diff --git a/.asf.yaml b/.asf.yaml new file mode 100644 index 000000000..befc9beb7 --- /dev/null +++ b/.asf.yaml @@ -0,0 +1,45 @@ +github: + description: "Community Over Code Asia - Asia edition of the flagship ASF conference" + homepage: https://asia.communityovercode.org + labels: + - apache + - community-over-code + - conference + features: + # Enable issues management + issues: true + # Enable projects for project management boards + projects: false + # Disable wiki for documentation + wiki: false + + enabled_merge_buttons: + # enable squash button: + squash: true + # default commit message when merging with a squash commit + # can either be: DEFAULT | PR_TITLE | PR_TITLE_AND_COMMIT_DETAILS | PR_TITLE_AND_DESC + squash_commit_message: PR_TITLE + # disable merge button: + merge: false + # default commit message when merging with a merge commit + # can either be: DEFAULT | PR_TITLE | PR_TITLE_AND_DESC + merge_commit_message: PR_TITLE_AND_DESC + # disable rebase button: + rebase: false + + rulesets: + - name: "Default Branch Protection" + type: branch + branches: + includes: + - "~DEFAULT_BRANCH" + - "release/*" + - "rel/*" + excludes: [] + bypass_teams: + - root + restrict_deletion: true + restrict_force_push: true +staging: + profile: acasia + whoami: gh-pages diff --git a/.github/dependabot.yml b/.github/dependabot.yml new file mode 100644 index 000000000..b5bb733bf --- /dev/null +++ b/.github/dependabot.yml @@ -0,0 +1,24 @@ +# Licensed to the Apache Software Foundation (ASF) under one +# or more contributor license agreements. See the NOTICE file +# distributed with this work for additional information +# regarding copyright ownership. The ASF licenses this file +# to you under the Apache License, Version 2.0 (the +# "License"); you may not use this file except in compliance +# with the License. You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, +# software distributed under the License is distributed on an +# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +# KIND, either express or implied. See the License for the +# specific language governing permissions and limitations +# under the License. + +version: 2 +updates: + - package-ecosystem: "github-actions" + directory: "/" + schedule: + interval: "monthly" + open-pull-requests-limit: 10 diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml new file mode 100644 index 000000000..5a11d9bd2 --- /dev/null +++ b/.github/workflows/build.yml @@ -0,0 +1,73 @@ +# Licensed to the Apache Software Foundation (ASF) under one +# or more contributor license agreements. See the NOTICE file +# distributed with this work for additional information +# regarding copyright ownership. The ASF licenses this file +# to you under the Apache License, Version 2.0 (the +# "License"); you may not use this file except in compliance +# with the License. You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, +# software distributed under the License is distributed on an +# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +# KIND, either express or implied. See the License for the +# specific language governing permissions and limitations +# under the License. + +name: Build docs + +on: + push: + branches: ['master'] + pull_request: + +jobs: + build: + runs-on: ubuntu-latest + + steps: + - name: 🗂 Checkout + uses: actions/checkout@v4 + with: + submodules: recursive + lfs: true + fetch-depth: 1 + - name: 📚 Install Hugo + uses: peaceiris/actions-hugo@16361eb4acea8698b220b76c0d4e84e1fd22c61d # v2.6.0 + with: + hugo-version: '0.159.1' # need to use this version to build the website + extended: true + - name: 🗂 Checkout gh-pages branch with minimum depth + uses: actions/checkout@v4 + if: ${{ github.event_name == 'push' }} + with: + ref: gh-pages + fetch-depth: 1 + path: dist + - name: 🔧 Build site + run: | + env HUGO_ENV="production" hugo -F + - name: 👷 Cleanup dist directory except .git, 2024,2025 and CNAME (each fork can have its own CNAME) + if: ${{ github.event_name == 'push' }} + run: find ./dist -mindepth 1 ! -path ./dist/CNAME ! -path './dist/.git*' ! -path './dist/202*' -delete -print + - name: 👷 Copy ./public/* and ./resources/* and .asf.yaml to /dist/ + if: ${{ github.event_name == 'push' }} + run: | + cp -r ./public/* ./dist/ + cp -r ./resources/* ./dist/ + # The asf.yaml file must be in the branch from which the files are published. + # Otherwise, ASF publishing tools cannot detect it. + cp .asf.yaml ./dist/ + + - name: 🚀 Add commit to the gh-pages branch and deploy it + if: ${{ github.event_name == 'push' }} + working-directory: dist + run: | + git config user.name "GitHub Actions" + git config user.email "actions@users.noreply.github.com" + git add . + git diff-index --quiet HEAD || git commit -m "Regenerated website based on ${GITHUB_SHA} commit" + git push origin gh-pages + + diff --git a/.gitignore b/.gitignore index bec26822d..ce2036a28 100644 --- a/.gitignore +++ b/.gitignore @@ -1,2 +1,12 @@ public datagen +docs +.DS_Store +resources/_gen +.idea/ +.hugo_build.lock +.vscode/ +script/env +node_modules/ +**/*.xlsx + diff --git a/README.md b/README.md new file mode 100644 index 000000000..489a91505 --- /dev/null +++ b/README.md @@ -0,0 +1,103 @@ +# ApacheCon hugo site + +This repository contains a Hugo based site for ApacheCon. + +NOTE: Please use **hugo v0.92.2** to avoid the website build error. + +To install Hugo : + +``` +brew install hugo +# or +port install hugo +``` + +Please chose hugo below v0.93, otherwise the build will failed. + +To test it locally, run: + +``` +hugo server -b http://localhost:1313/acasia2023 -F +``` + +To generate final site, use: + + +``` +hugo -b https://apachecon.com/acasia2023/ -d -F +``` + +## FAQ + +### How to fix the typos in Session page +If you want to fix session typos of https://asia.communityovercode.org/sessions/cloudnative-913295.html +You just need to go the file cloudnative-913295.md in content/sessions directory. +If it is chinese version, you just need to update the file of cloudnative-913295.zh.md. + + +### How can I publish the website to ApacheCon +First you need to write access right of [ApacheCon website] + +Checking out the this repo +``` +git@github.com:apache/acasia-website.git +``` + +After doing some modifications on the markdown files, you can use hugo command to generate the website from the source directory +``` +cd +hugo -b https://asia.communityovercode.org/ -d -F +``` + + +### How can I modify the content of the main page: + +It's under `content/_index.md` + +### How can I have translations: + +Add new language to the `config.toml`: + +``` +[languages.zh] +languageName = "中文" +weight = 2 +``` + +And create language specific version for each content file: + +``` +content/_index.md +content/_index.zh.md +content/cfp.md +content/cfp.zh.md +``` + +### How can I modify the menu? + +For simple links, add the menu entry to the `config.toml`: + +(Note: add it for all the active languages) + +``` +[[languages.en.menu.main]] +name = "t-shirt" +url = "https://s.apache.org/apache-tshirt" +weight = -110 +``` + +The most page links of the menu are generated from the markdown files of content directory. +For a page, it's enough to add a menu entry to the markdown (For example `content/cfp.md`): + +``` +--- +title: Call for Presentations +menu: + main: + weight: -300 +--- +... +... +``` + +You can add or remove the links from navigation bar by adding or removing this menu entry. diff --git a/assets/scss/acasia2021.scss b/assets/scss/acasia2021.scss new file mode 100644 index 000000000..c12bf9034 --- /dev/null +++ b/assets/scss/acasia2021.scss @@ -0,0 +1,67 @@ +/* acasia2021.scss */ +section.session { + border-bottom: #1aa7e3 solid 1px; + article { + padding-bottom: 20px; + } + dl { + padding-top: 10px; + overflow: hidden; + height: auto; + border-top: #1aa7e3 solid 1px; + dt { + font-weight: normal; + vertical-align: middle; + } + dt,dd { + float: left; + display: inline-block; + } + a { + color: #1aa7e3; + } + h3 { + font-weight: normal; + } + } + + .session-date { + width: 12%; + } + .session-list { + width: 85%; + overflow: hidden; + } + + .top-30 { + padding-top: 30px; + } + .lh-2em { + line-height: 2em; + } +} + + +section.sponsorship{ + text-align: center; + overflow: hidden; + dl{ + padding-top: 2em; + overflow: auto; + dt{ + padding-bottom: 1em; + font-size: 2em; + } + dd { + overflow: auto; + min-height: 100px; + line-height: 100px; + vertical-align: baseline; + img { + min-width: 100px; + max-width: 300px; + padding: 0 10px; + } + } + } +} diff --git a/assets/scss/main.scss b/assets/scss/main.scss new file mode 100644 index 000000000..119c0193f --- /dev/null +++ b/assets/scss/main.scss @@ -0,0 +1,28 @@ +article { + width: 100%; + h3 { + margin-bottom: 1rem; + } + pre { + background-color: #e9ecef; + border-radius: 0.25rem; + padding: 1rem; + } + blockquote { + p { + color: var(--bs-body-color); + } + > p:last-child { + margin-bottom: 0; + } + } + table { + td, + th { + word-break: initial; + } + } + ol ol,ol ul,ul ol,ul ul { + margin-bottom: 1rem; + } +} \ No newline at end of file diff --git a/assets/scss/schedule.scss b/assets/scss/schedule.scss new file mode 100644 index 000000000..336a0d10d --- /dev/null +++ b/assets/scss/schedule.scss @@ -0,0 +1,701 @@ +/* schedule.scss - session grid on /sessions */ + +$schedule-ink: #0f072d; +$schedule-line: #dcdfe6; +$schedule-accent: #EA7826; + +.schedule-tabs { + display: flex; + flex-wrap: wrap; + gap: 6px; + border-bottom: 1px solid $schedule-ink; + margin-bottom: 20px; + + .schedule-tab { + position: relative; + bottom: -1px; + margin: 0; + padding: 6px 16px; + font-size: 18px; + line-height: 1.4; + background-color: transparent; + border: 1px solid $schedule-ink; + border-bottom-color: transparent; + border-radius: 4px 4px 0 0; + cursor: pointer; + + &.is-selected { + background-color: $schedule-ink; + color: #fff; + } + } +} + +.schedule-summary { + display: flex; + flex-wrap: wrap; + align-items: baseline; + gap: 12px; + margin-bottom: 12px; + font-size: 15px; + color: #555; + + .schedule-date { + font-size: 22px; + font-weight: 700; + color: $schedule-ink; + } + + .schedule-count { + color: $schedule-accent; + } + + .schedule-hint { + display: none; + margin-left: auto; + font-size: 13px; + } +} + +.schedule-scroll { + overflow-x: auto; + -webkit-overflow-scrolling: touch; + border: 1px solid $schedule-line; + border-radius: 6px; + max-height: 80vh; + overflow-y: auto; +} + +.schedule-block + .schedule-block { + margin-top: 18px; +} + +/* a block with a single room (plenary / keynote) fills the width */ +.schedule-block.is-single-room { + .schedule-grid { + width: 100%; + } + + .col-room { + min-width: 0; + } +} + +.schedule-grid { + width: auto; + min-width: 100%; + border-collapse: separate; + border-spacing: 0; + font-size: 14px; + background: #fff; + + th, + td { + border-right: 1px solid $schedule-line; + border-bottom: 1px solid $schedule-line; + vertical-align: top; + padding: 8px; + + &:last-child { + border-right: 0; + } + } + + thead th { + position: sticky; + top: 0; + z-index: 3; + background: #f6f7f9; + text-align: left; + font-weight: 700; + color: $schedule-ink; + white-space: normal; + } + + .col-room { + min-width: 230px; + + .room-name { + display: block; + font-size: 15px; + } + + .room-tracks { + display: flex; + flex-wrap: wrap; + gap: 4px; + margin-top: 4px; + } + + .room-track { + font-size: 11px; + font-weight: 600; + padding: 1px 6px; + border-radius: 10px; + text-decoration: none; + color: hsl(var(--sess-hue), 55%, 28%); + background: hsl(var(--sess-hue), 60%, 92%); + + &:hover { + background: hsl(var(--sess-hue), 60%, 85%); + } + } + } + + .col-time { + position: sticky; + left: 0; + z-index: 2; + min-width: 84px; + background: #f6f7f9; + font-weight: 700; + color: $schedule-ink; + text-align: left; + white-space: nowrap; + } + + thead .col-time { + z-index: 4; + } + + .tz { + display: block; + font-size: 11px; + font-weight: 400; + color: #888; + } + + tbody tr:last-child th, + tbody tr:last-child td { + border-bottom: 0; + } + + tbody tr:hover .cell { + background: #fcfcfd; + } + + .cell.is-empty { + background: repeating-linear-gradient(135deg, #fafafa, #fafafa 6px, #f3f4f6 6px, #f3f4f6 12px); + } +} + +.sess { + padding: 7px 10px; + border-left: 4px solid hsl(var(--sess-hue), 60%, 45%); + border-radius: 4px; + background: hsl(var(--sess-hue), 65%, 97%); + + & + .sess { + margin-top: 6px; + } + + .sess-title { + display: block; + font-weight: 600; + font-size: 14px; + line-height: 1.35; + color: $schedule-ink; + text-decoration: none; + + &:hover { + color: hsl(var(--sess-hue), 60%, 38%); + text-decoration: underline; + } + } + + .sess-speaker { + margin-top: 3px; + font-size: 12px; + color: #5a5a6a; + } + + .sess-meta { + display: flex; + flex-wrap: wrap; + align-items: center; + gap: 5px; + margin-top: 5px; + font-size: 11px; + } + + .sess-track { + font-weight: 600; + padding: 1px 6px; + border-radius: 10px; + text-decoration: none; + color: hsl(var(--sess-hue), 55%, 28%); + background: hsl(var(--sess-hue), 60%, 90%); + } + + .sess-room, + .sess-type { + color: #6b6b7b; + } + + .sess-room::before { + content: "\2302\00a0"; + } +} + +/* stacked view for narrow screens */ +.schedule-list { + display: none; + + &.is-visible { + display: block; + } + + .slot { + padding: 10px 0; + border-bottom: 1px solid $schedule-line; + + &:last-child { + border-bottom: 0; + } + } + + .slot-time { + font-size: 18px; + font-weight: 700; + color: $schedule-ink; + margin-bottom: 8px; + + .tz { + font-size: 12px; + font-weight: 400; + color: #888; + } + } + + .slot-sessions .sess { + margin-bottom: 6px; + } +} + +@media screen and (max-width: 860px) { + .schedule-scroll { + display: none; + } + + .schedule-list { + display: block; + } + + .schedule-tabs .schedule-tab { + font-size: 16px; + padding: 6px 10px; + } +} + +@media screen and (min-width: 861px) { + .schedule-summary .schedule-hint { + display: inline; + } +} + +/* ------------------------------------------------------------------ * + * Track detail page (/tracks/.html) + * ------------------------------------------------------------------ */ + +.track-page { + .track-hero { + padding: 18px 22px; + margin-bottom: 26px; + border-left: 5px solid hsl(var(--track-hue), 60%, 45%); + border-radius: 6px; + background: linear-gradient(90deg, hsl(var(--track-hue), 65%, 96%), #fff 85%); + } + + .track-back { + display: inline-block; + font-size: 13px; + font-weight: 600; + text-decoration: none; + color: hsl(var(--track-hue), 55%, 32%); + + &:hover { + text-decoration: underline; + } + } + + .track-title { + margin: 6px 0 14px; + font-size: 34px; + line-height: 1.25; + color: $schedule-ink; + } + + .track-facts { + display: flex; + flex-wrap: wrap; + gap: 10px 32px; + margin: 0; + + .fact { + display: flex; + flex-direction: column; + gap: 3px; + } + + dt { + font-size: 11px; + font-weight: 700; + letter-spacing: 0.08em; + text-transform: uppercase; + color: #7a7a8a; + } + + dd { + margin: 0; + display: flex; + flex-wrap: wrap; + gap: 6px; + font-size: 15px; + color: $schedule-ink; + } + + .chair { + padding: 1px 8px; + border-radius: 10px; + font-size: 13px; + background: hsl(var(--track-hue), 60%, 92%); + color: hsl(var(--track-hue), 55%, 28%); + } + } + + .track-intro { + max-width: 900px; + margin-bottom: 34px; + font-size: 15px; + line-height: 1.75; + + p { + margin-bottom: 1em; + } + } + + .track-agenda-title { + margin: 0 0 14px; + padding-bottom: 8px; + font-size: 22px; + border-bottom: 2px solid hsl(var(--track-hue), 60%, 45%); + color: $schedule-ink; + } + + .track-day + .track-day { + margin-top: 22px; + } + + .track-day-head { + display: flex; + flex-wrap: wrap; + align-items: baseline; + justify-content: space-between; + gap: 8px; + margin-bottom: 8px; + padding: 5px 10px; + border-radius: 4px; + background: hsl(var(--track-hue), 45%, 95%); + + .track-day-date { + font-size: 17px; + font-weight: 700; + color: $schedule-ink; + } + + .track-day-link { + font-size: 12px; + text-decoration: none; + color: hsl(var(--track-hue), 55%, 35%); + + &:hover { + text-decoration: underline; + } + } + } + + .track-slots { + list-style: none; + margin: 0; + padding: 0; + } + + /* one session per row: time in the left gutter, speaker and room to the right */ + .track-slot { + display: flex; + gap: 12px; + padding: 0; + + & + .track-slot { + margin-top: 6px; + } + + .track-slot-time { + flex: 0 0 62px; + padding-top: 9px; + font-size: 14px; + font-weight: 700; + color: $schedule-ink; + + .tz { + display: block; + font-size: 10px; + font-weight: 400; + color: #999; + } + } + + .sess { + flex: 1 1 auto; + min-width: 0; + display: flex; + flex-wrap: wrap; + align-items: baseline; + gap: 4px 16px; + padding: 9px 12px; + } + + .sess-title { + flex: 1 1 340px; + font-size: 15px; + } + + .sess-speaker, + .sess-meta { + flex: 0 0 auto; + margin-top: 0; + font-size: 12px; + } + } + + .track-tba { + padding: 14px; + border: 1px dashed $schedule-line; + border-radius: 6px; + color: #6b6b7b; + } +} + +@media screen and (max-width: 980px) { + .track-page { + .track-title { + font-size: 27px; + } + + .track-intro { + margin-bottom: 26px; + } + + /* stacked card again: a single row per session gets too cramped */ + .track-slot .sess { + display: block; + } + + .track-slot .sess-speaker { + margin-top: 3px; + } + + .track-slot .sess-meta { + margin-top: 5px; + display: flex; + flex-wrap: wrap; + } + } +} + +/* ------------------------------------------------------------------ * + * Session detail page (/sessions/.html) + * ------------------------------------------------------------------ */ + +.session-detail { + max-width: 860px; + margin: 0 auto; + + .session-hero { + padding: 20px 24px; + margin-bottom: 26px; + border-left: 5px solid hsl(var(--track-hue), 60%, 45%); + border-radius: 6px; + background: linear-gradient(90deg, hsl(var(--track-hue), 65%, 96%), #fff 88%); + } + + .session-crumbs { + font-size: 13px; + margin-bottom: 10px; + color: #9a9aa8; + + a { + color: hsl(var(--track-hue), 55%, 32%); + text-decoration: none; + + &:hover { text-decoration: underline; } + } + + .sep { margin: 0 6px; } + } + + .session-title { + margin: 0 0 12px; + font-size: 32px; + line-height: 1.25; + color: $schedule-ink; + } + + .session-presenters { + display: flex; + flex-wrap: wrap; + gap: 6px; + margin-bottom: 16px; + + .presenter { + font-size: 14px; + font-weight: 600; + padding: 2px 10px; + border-radius: 12px; + color: hsl(var(--track-hue), 55%, 28%); + background: hsl(var(--track-hue), 60%, 92%); + } + } + + .session-meta { + display: flex; + flex-wrap: wrap; + gap: 8px 20px; + margin: 0; + padding: 0; + list-style: none; + font-size: 14px; + color: #55555f; + + li { + display: flex; + align-items: center; + gap: 6px; + } + + .ico { + font-size: 15px; + line-height: 1; + filter: grayscale(0.2); + } + + .tz { + margin-left: 3px; + font-size: 11px; + color: #999; + } + + .meta-track a { + font-weight: 600; + text-decoration: none; + color: hsl(var(--track-hue), 55%, 32%); + + &:hover { text-decoration: underline; } + } + } + + .session-body { + font-size: 16px; + line-height: 1.8; + color: #2c2c35; + + p { margin-bottom: 1.1em; } + + /* the "Speakers" / 讲师 heading that precedes the bios */ + h3 { + margin: 30px 0 16px; + padding-bottom: 8px; + font-size: 20px; + border-bottom: 2px solid hsl(var(--track-hue), 60%, 45%); + color: $schedule-ink; + } + + img { + float: left; + width: 130px !important; + height: auto; + margin: 4px 20px 12px 0; + border-radius: 8px; + box-shadow: 0 2px 10px rgba(0, 0, 0, 0.12); + } + + /* clear the floated portrait before the next block */ + h3 + p, + br + img { clear: left; } + } + + .session-video { + position: relative; + padding-bottom: 56.25%; + height: 0; + margin: 28px 0; + border-radius: 8px; + overflow: hidden; + + iframe { + position: absolute; + top: 0; + left: 0; + width: 100%; + height: 100%; + } + } + + .session-nav { + display: flex; + gap: 16px; + margin-top: 36px; + padding-top: 20px; + border-top: 1px solid $schedule-line; + + .nav-prev, + .nav-next { + flex: 1 1 0; + min-width: 0; + } + + .nav-next { text-align: right; } + + a { + display: inline-flex; + flex-direction: column; + gap: 3px; + text-decoration: none; + color: $schedule-ink; + + &:hover .nav-title { color: hsl(var(--track-hue), 55%, 38%); } + } + + .dir { + font-size: 12px; + font-weight: 600; + color: hsl(var(--track-hue), 55%, 35%); + } + + .nav-title { + font-size: 14px; + display: -webkit-box; + -webkit-line-clamp: 2; + -webkit-box-orient: vertical; + overflow: hidden; + } + } +} + +@media screen and (max-width: 640px) { + .session-detail { + .session-title { font-size: 25px; } + + .session-body img { + float: none; + display: block; + margin: 4px 0 12px; + } + + .session-nav { + flex-direction: column; + + .nav-next { text-align: left; } + } + } +} diff --git a/config.toml b/config.toml index 192e5b110..7a4d0f5a9 100644 --- a/config.toml +++ b/config.toml @@ -1,6 +1,75 @@ -baseURL = "http://example.org/" -languageCode = "en-us" -title = "ApacheCon at Home 2021" +# set the baseURL for CI verification +baseURL = "/" +title = "Community Over Code Asia 2026" theme = "apachecon" # When enabled, creates URL of the form /filename.html instead of /filename/. -uglyURLs = true \ No newline at end of file +uglyURLs = true +#hasCJKLanguage = true +enableRobotsTXT = true + +languageCode = "en-us" +defaultContentLanguage = "en" + + + +[params] + author = "ALC-Beijing" + description = "Welcome to Community Over Code (ApacheCon Asia) 2026" + keywords = ["Community Over Code","Community Over Code Asia","ApacheCon Asia 2026", "Apache 软件基金会亚洲大会 2026"] + schedule = ["2026-08-07", "2026-08-08", "2026-08-09"] + +[languages.en] +languageName = "English" +weight = 1 + +[languages.zh] +languageName = "中文" +weight = 2 + +disableKinds = ["taxonomy", "taxonomyTerm"] + +[markup.goldmark.renderer] +unsafe = true + +# menu for en (final list includes translated pages from /content) +[[languages.en.menu.main]] +name = "Tracks" +url = "/tracks" +weight = -100 + +[[languages.en.menu.main]] +name = "Sessions" +url = "/sessions" +weight = -50 + +[[languages.en.menu.main]] +name = "Guide" +url = "/guide/visa_letter_request.html" +weight = 4 + +[[languages.en.menu.main]] +name = "Venue + Travel" +url = "/venue_travel/venue.html" +weight = 5 + +# menu for ZH language (final list includes translated pages from /content) +[[languages.zh.menu.main]] +name = "议题" +url = "/zh/sessions" +weight = -50 + +# menu for ZH language (final list includes translated pages from /content) +[[languages.zh.menu.main]] +name = "专题" +url = "/zh/tracks" +weight = -100 + +[[languages.zh.menu.main]] +name = "参会指南" +url = "/zh/guide/visa_letter_request.html" +weight = 4 + +[[languages.zh.menu.main]] +name = "会场 + 旅行" +url = "/zh/venue_travel/venue.html" +weight = 5 \ No newline at end of file diff --git a/content/_index.md b/content/_index.md new file mode 100644 index 000000000..31e69ba27 --- /dev/null +++ b/content/_index.md @@ -0,0 +1,13 @@ +--- +layout: home +title: Community Over Code Asia 2026 will be held in Beijing from August 7 to 9, 2026 +keywords: Community Over Code Asia, Community Over Code, ApacheCon, Community Over Code Asia 2026 +--- + +Community Over Code (formerly known as ApacheCon) is the official global conference series of The Apache Software Foundation (ASF). Since 1998 – before the ASF’s incorporation – ApacheCon has been drawing participants at all levels to explore ”Tomorrow’s Technology Today” across 350+ Apache projects and their diverse communities. Community Over Code showcases the latest developments in Apache projects and emerging innovations through hands-on sessions, keynotes, real-world case studies, trainings, hackathons, and more. + + +Community Over Code showcases the latest breakthroughs from ubiquitous Apache projects and upcoming innovations in the Apache Incubator, as well as open source development and leading community-driven projects the Apache way. Attendees learn about core open source technologies independent of business interests, corporate biases, or sales pitches. + + +The Community Over Code program is dynamic, evolving at each event with content directly driven by select Apache project developer and user communities. Community Over Code delivers state-of-the-art content that features the latest open source advances in big data, cloud, community development, FinTech, IoT, machine learning, messaging, programming, search, security, servers, streaming, web frameworks, and more in a collaborative, vendor-neutral environment. diff --git a/content/_index.zh.md b/content/_index.zh.md new file mode 100644 index 000000000..a8c13b561 --- /dev/null +++ b/content/_index.zh.md @@ -0,0 +1,10 @@ +--- +title: Community Over Code Asia 2026 - ASF旗下大会 将于2026年8月7-9日在北京举行 +keywords: Community Over Code Asia, Community Over Code, ApacheCon, Community Over Code Asia 2026 +--- + +Community Over Code 是 Apache 软件基金会（ASF）的官方全球系列大会，其前身为 ApacheCon。自 1998 年以来，在 ASF 成立之前，ApacheCon 已经吸引了各个层次的参与者，在 350 多个 Apache 项目及其不同的社区中探索 “明天的技术”。Community Over Code 通过动手实作、主题演讲、实际案例研究、培训、黑客松活动等方式，展示 Apache 项目的最新发展和新兴创新。 + +Community Over Code 展示了无处不在的 Apache 项目的最新突破和 Apache 孵化器中即将到来的创新，以及开源开发和以 Apache 之道领导社区驱动的项目。与会者可以了解到独立于商业利益、企业偏见或推销话术之外的核心开源技术。 + +Community Over Code 项目是动态的，每次活动的内容都是由精选的 Apache 项目开发者和用户社区直接推动的。Community Over Code 提供了最先进的内容，在一个协作、厂商中立的环境中，展示了大数据、云计算、社区发展、金融科技、物联网、机器学习、消息中间件、编程、搜索、安全、服务器、流媒体、网络框架等方面的最新开源进展。 diff --git a/content/cfp.md b/content/cfp.md deleted file mode 100644 index d2da02269..000000000 --- a/content/cfp.md +++ /dev/null @@ -1,38 +0,0 @@ ---- - -title: Call for Presentations ---- -The call for presentations is now open! Submit proposals by **Monday, May 3rd, 2021 8:00** AM (America/New_York time - UTC -5). **Please do not wait until the last minute.** - -[Submit your talk proposal now!](https://acah2021.jamhosted.net/) - -We are looking for presentations about anything relating to Apache Software Foundation projects, open source governance, community, and software development. - -In particular, this year we are building content tracks around the following specific topics/projects: - - * API/Microservice - * Cassandra - * Content Delivery - * Community - * Drill - * Geospatial - * Incubating - * IoT - * Observability - * Royale - * Tomcat - * Big Data - * Community - * Content Management - * Fineract/Fintech - * Groovy - * Integration - * Karaf - * RDF/Linked Data - * Search - * Social Data - -If your proposed presentation falls into one of these categories, please select that topic in the CfP entry form. Or select **General** if it's related to some other topic or project area. - -[Submit your talk proposal now!](https://acah2021.jamhosted.net/) - diff --git a/content/conduct.md b/content/conduct.md deleted file mode 100644 index b670e7dbe..000000000 --- a/content/conduct.md +++ /dev/null @@ -1,16 +0,0 @@ ---- -title: ApacheCon Code of Conduct ---- -At the Apache Software Foundation, all online interactions are governed by the [ASF Code of Conduct](http://apache.org/foundation/policies/conduct.html), while in-person events are subject to the [Anti-Harassment Policy](http://apache.org/foundation/policies/anti-harassment.html). - -ApacheCon is dedicated to providing a harassment-free experience for everyone. We do not tolerate harassment of participants in any form. Participants violating these rules may be sanctioned or expelled without a refund, at the discretion of the event organizers. - -Harassment includes offensive verbal comments, deliberate intimidation, stalking, following, unwanted photography or recording, sustained disruption of talks or other events, inappropriate physical contact, and unwelcome sexual attention. Participants asked to stop any harassing behaviour are expected to comply immediately. - -Sexual language and imagery will not be tolerated in any event venue, including talks. Exhibitors should also refrain from using sexualised images, activities, or other material. Booth staff (including volunteers) should not use sexualised clothing/uniforms/costumes, or otherwise create a sexualised environment. - -If a participant engages in harassing behaviour, the organizers may take any action they deem appropriate, including warning the offender or expulsion from the event without a refund. We expect participants to follow these rules at all event venues and related social events. - -If you are being harassed, notice that someone else is being harassed, or have any other concerns, please contact a member of the event team immediately. The team can be found at the registration desk. - -You can contact the event team via email at [planners@apachecon.com](planners@apachecon.com), or via the live chat feature on the event website. \ No newline at end of file diff --git a/content/guide/accessibility.md b/content/guide/accessibility.md new file mode 100644 index 000000000..8bf5f07ad --- /dev/null +++ b/content/guide/accessibility.md @@ -0,0 +1,8 @@ +--- +title: Accessiblity +--- + + +Community Over Code Asia strives to be a welcoming, inclusive, and accessible environment for all, regardless of ability. + +Participants requiring special accommodations should provide details in their registration and/or by emailing us at [ningjiang@apache.org](mailto:ningjiang@apache.org). This information will be kept private and only be shared with individuals who need to know to secure these accommodations. \ No newline at end of file diff --git a/content/guide/accessibility.zh.md b/content/guide/accessibility.zh.md new file mode 100644 index 000000000..6a0c79620 --- /dev/null +++ b/content/guide/accessibility.zh.md @@ -0,0 +1,7 @@ +--- +title: 无障碍设施 +--- + +Community Over Code Asia 力求为所有人提供一个欢迎、包容和无障碍的环境，无论其能力如何。 + +需要特殊配套服务的参会者请在注册时提供详细信息，并发送电子邮件至 [ningjiang@apache.org](mailto:ningjiang@apache.org) 与我们联系。您的相关信息将被保密，只有需要知道以确保提供这些配套服务的人员才能了解。 \ No newline at end of file diff --git a/content/guide/code_of_conduct.md b/content/guide/code_of_conduct.md new file mode 100644 index 000000000..9b2bc7783 --- /dev/null +++ b/content/guide/code_of_conduct.md @@ -0,0 +1,17 @@ +--- +title: Code of conduct +weight: 22 +--- + +At the Apache Software Foundation, all online interactions are governed by the ASF [Code of Conduct](https://apache.org/foundation/policies/conduct.html), while in-person events are subject to the [Anti-Harassment Policy](https://apache.org/foundation/policies/anti-harassment.html). + +Community Over Code (formerly ApacheCon) is dedicated to providing a harassment-free experience for everyone. We do not tolerate harassment of participants in any form. Participants violating these rules may be sanctioned or expelled without a refund, at the discretion of the event organizers. + +Harassment includes offensive verbal comments, deliberate intimidation, stalking, following, unwanted photography or recording, sustained disruption of talks or other events, inappropriate physical contact, and unwelcome sexual attention. Participants asked to stop any harassing behaviour are expected to comply immediately. + +Sexual language and imagery will not be tolerated in any event venue, including talks. Exhibitors should also refrain from using sexualised images, activities, or other material. Booth staff (including volunteers) should not use sexualised clothing/uniforms/costumes, or otherwise create a sexualised environment. +If a participant engages in harassing behaviour, the organizers may take any action they deem appropriate, including warning the offender or expulsion from the event without a refund. We expect participants to follow these rules at all event venues and related social events. + +If you are being harassed, notice that someone else is being harassed, or have any other concerns, please contact a member of the event team immediately. The team can be found at the registration desk. + +You can contact the event team via email at [planners@apachecon.com](mailto:planners@apachecon.com). \ No newline at end of file diff --git a/content/guide/code_of_conduct.zh.md b/content/guide/code_of_conduct.zh.md new file mode 100644 index 000000000..b32a72b4d --- /dev/null +++ b/content/guide/code_of_conduct.zh.md @@ -0,0 +1,18 @@ +--- +title: 行为准则 +weight: 22 +--- + +在Apache软件基金会，所有在线互动均受[ASF行为准则](http://apache.org/foundation/policies/conduct.html)的管辖，而面对面的活动则受[反骚扰政策](http://apache.org/foundation/policies/anti-harassment.html)。 + +Community Over Code (前 ApacheCon ) 致力于为每个人提供无骚扰的体验。我们不容忍任何形式的参与者骚扰。违反这些规则的参加者可以根据活动组织者的判断，予以制裁或开除，且不予退款。 + +骚扰包括令人反感的口头评论，故意的恐吓，跟踪，跟踪，不必要的摄影或录音，谈话或其他事件的持续中断，不适当的身体接触以及不受欢迎的性关注。被要求停止任何骚扰行为的参与者应立即遵守。 + +在包括演讲在内的任何活动场所，都不会容忍性语言和图像。参展商还应避免使用带有色情色彩的图像，活动或其他材料。展位人员（包括志愿者）不得使用带有性服装/制服/服装，或以其他方式营造性环境。 + +如果参与者进行骚扰行为，组织者可以采取他们认为适当的任何措施，包括警告违法者或开除该事件，不予退款。我们希望参与者在所有活动场所和相关社交活动中都遵守这些规则。 + +如果您受到骚扰，请注意有人受到骚扰，或有其他任何疑虑，请立即与活动团队成员联系。该团队可以在注册处找到。 + +您可以通过电子邮件 [planners@apachecon.com](mailto:planners@apachecon.com)或活动网站上的实时聊天功能与活动团队联系。 diff --git a/content/guide/convince_your_boss.md b/content/guide/convince_your_boss.md new file mode 100644 index 000000000..62177507d --- /dev/null +++ b/content/guide/convince_your_boss.md @@ -0,0 +1,10 @@ +--- +title: Convince Your Boss +weight: 22 +--- + +Want to attend but not sure how to bring it up with your boss? We’ve created this sample email to help you explain why it’s important to attend Community Over Code Asia 2025. Feel free to use the full letter or pieces that work for you. See you in Beijing! + + +Download Template + \ No newline at end of file diff --git a/content/guide/convince_your_boss.zh.md b/content/guide/convince_your_boss.zh.md new file mode 100644 index 000000000..8464030b6 --- /dev/null +++ b/content/guide/convince_your_boss.zh.md @@ -0,0 +1,8 @@ +--- +title: 向您所在公司阐述参会价值 +weight: 22 +--- + +想参加本届大会但不确定如何向老板提出申请？我们准备了这份邮件模板来帮助您向公司解释为何参加 Community Over Code Asia 很重要。您可以视需要使用整封邮件或适合您的部分。我们期待在北京与您相聚！ + +下载模板 \ No newline at end of file diff --git a/content/guide/presentation_slides_template.md b/content/guide/presentation_slides_template.md new file mode 100644 index 000000000..160a1dd9c --- /dev/null +++ b/content/guide/presentation_slides_template.md @@ -0,0 +1,17 @@ +--- +title: Make a presentation slide +weight: 21 +--- + +We have a [slides template](https://docs.google.com/presentation/d/1cAZfwR5Rs8pq6SCkguT-imlSpPVMKNpng1nJnjJe0MU/edit?usp=sharing). + +**We strongly recommend you to use the template, but if you don’t plan to use it for your entire presentation, please use the title slide at least.** + +- If you use your own template, please set your slide size ratio as 16:9. +- Do not include any offensive or sexualized references. This would be a violation of the Code of Conduct. + +**Talks are 30 minutes long. This time is inclusive of Q&A.** + +The ideal time structure of your presentation would be +- 23-27 minutes presentation. +- 3-5 min Q&A. \ No newline at end of file diff --git a/content/guide/presentation_slides_template.zh.md b/content/guide/presentation_slides_template.zh.md new file mode 100644 index 000000000..b7d8550f4 --- /dev/null +++ b/content/guide/presentation_slides_template.zh.md @@ -0,0 +1,18 @@ +--- +title: 制作演示文档 +weight: 21 +--- + +您可以在此处下载我们的 PPT 模板 [slides template](https://docs.google.com/presentation/d/1cAZfwR5Rs8pq6SCkguT-imlSpPVMKNpng1nJnjJe0MU/edit?usp=sharing). 如果您所在公司无法访问 Google Docs，请联系 [annetta.li@csdn.net](mailto:annetta.li@csdn.net)，我们将邮件为您发送该文件。 + +**我们强烈建议您使用大会统一的 PPT 模板，但如果您不计划使用该模板，请至少在封面封底页使用同一的模板**。 + +- 如果您使用自己的 PPT 模板，请将幻灯片大小比例设置为 16:9。 +- 请勿包含任何攻击性或色情方面的内容，这将违反《行为准则》。 + +**演讲时长为 30 分钟。这个时间包括 Q&A**。 + +理想的演讲时间分配为： + +- 23-27 分钟演讲； +- 3-5 分钟问答。 \ No newline at end of file diff --git a/content/guide/terms.md b/content/guide/terms.md new file mode 100644 index 000000000..bf929d906 --- /dev/null +++ b/content/guide/terms.md @@ -0,0 +1,19 @@ +--- +title: Terms & Conditions +--- + +## Refund Policy + +If you must cancel for any reason please notify us via email at [vivian@segmentfault.com](mailto:vivian@segmentfault.com). Refunds will only be issued for cancellations received 14 days prior to the first day of the event and may take up to 14 days to process. Please note that all refunds are subject to a 3% processing fee. You can cancel or modify your registration from your confirmation email. + +## Photo & Video Release + +I hereby consent to the photographing of myself and the recording of my voice and the use of these photographs and/or recordings singularly or in conjunction with other photographs and/or recordings for advertising, publicity, commercial or other business purposes for The Apache Software Foundation. I understand that the term “photograph” as used herein encompasses both still photographs and motion picture footage. + +## GDPR/Data Protection + +By submitting this registration form, you understand that the data you include will be submitted to The Apache Software Foundation, and that The Apache Software Foundation may contact you regarding this event. Personal data collected from this registration form is subject to The Apache Software Foundation’s [Privacy Policy](https://privacy.apache.org/policies/privacy-policy-public.html). + +In order to facilitate networking and business relationships at the event, you may be asked by third parties at the event for permission to scan your attendee badge via a lead retrieval system. You are never required to have your badge scanned by third parties. Allowing your attendee badge to be scanned enables the scanner to receive certain of your registration data, including but not limited to your name, phone number, address, email, employer, job title and job function. If you choose to allow your attendee badge to be scanned, you are explicitly consenting to receipt and use of such data by the third-party recipients, which will be subject to their own privacy policies. + +By signing up to attend a Co-Located Event hosted by a third party, you understand that a portion of the data you submit will be provided to the third party, to enable the third party to hold the event. The third parties’ use of the data will be subject to their own privacy policies. diff --git a/content/guide/terms.zh.md b/content/guide/terms.zh.md new file mode 100644 index 000000000..58a429f07 --- /dev/null +++ b/content/guide/terms.zh.md @@ -0,0 +1,22 @@ +--- +title: 条款和条件 +--- + +## 退款政策 + +如果您因任何原因必须取消注册，请发送邮件至 [vivian@segmentfault.com](mailto:vivian@segmentfault.com) 通知我们。退款只针对在活动第一天的 14 天前收到的取消进行，并可能需要 14 天的处理时间。请注意，所有退款均会产生 3％的手续费。 +如果您无法参加，需要替换参与者，您可以在邮件里面编辑参会信息。 + +如果您无法参加，需要替换参与者，您可以在邮件里面编辑参会信息。 + +## 照片和视频发布 + +我在此同意，Apache 软件基金会有权出于广告、宣传、商业或其他业务目的对我本人进行拍照、对我的声音进行录制，并单独使用这些照片和/或录音，或将这些照片和/或录音与其他照片和/或录音一起使用。我同意此处所用的“照片”一词既包括静态照片，也包括动态影像片段。 + +## 《通用数据保护条例》/ 数据保护 + +通过提交此注册表，表明您已了解您输入的数据将被提交给Apache 软件基金会，并且Apache 软件基金会可以就此活动与您联系。从此注册表中收集的个人数据受 Apache 软件基金会的[隐私政策](https://privacy.apache.org/policies/privacy-policy-public.html)约束。 + +为了促进活动中的商务交流和业务关系，参与活动的第三方可能会要求通过客户检索系统扫描您的参会名牌。任何人绝不强制您同意第三方扫描你的名牌。若您允许第三方扫描您的参会名牌，您将使扫描仪能够接收您的某些注册数据，包括但不限于您的姓名、电话号码、地址、电子邮箱、雇主，职位和工作职能等。如果您选择允许扫描您的参会名牌，则您明确同意由第三方根据其隐私政策来接收和使用此类数据。 + +通过注册参加由第三方主办的同场活动，表明您已了解您提交的部分数据将被提供给第三方，以使第三方能够举办该活动。第三方对数据的使用将受其自身隐私政策的约束。 diff --git a/content/guide/visa_letter_request.md b/content/guide/visa_letter_request.md new file mode 100644 index 000000000..50d0a33d9 --- /dev/null +++ b/content/guide/visa_letter_request.md @@ -0,0 +1,43 @@ +--- +title: Visa Letter Request +weight: 20 +--- + +For those needing a visa invitation letter, please refer to the instructions on this page. + +### INSTRUCTIONS + +Before asking for a visa invitation letter, please check out [the website of Chinese Visa Application Service Center](https://www.visaforchina.cn/globle/) for the visa application in your city. + +### Please pay attention to the latest visa policy of China: + + - **240-Hour Visa-Free Transit Policy which allows people from certain countries to enjoy ten days of travel to select areas of the country without applying for a visa beforehand.** + + - To obtain this visa exemption, the foreign national must have a valid passport from one of the 54 countries, which includes: Austria, Belgium, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Poland, Portugal, Slovakia, Slovenia, Spain, Sweden, Switzerland, Russia, United Kingdom, Ireland, Cyprus, Bulgaria, Romania, Ukraine, Serbia, Croatia, Bosnia and Herzegovina, Montenegro, North Macedonia, Albania, Monaco, Belarus, Norway，The United States, Canada, Brazil, Mexico, Argentina, Chile, Australia, New Zealand, South Korea, Japan, Singapore, Brunei, the United Arab Emirates, and Qatar. + + - **As per the requirements of China’s National Immigration Authority (NIA), people applying for 240-hour visa-free transit must have:** + + - A valid passport or another international travel document that is valid for more than three months from the arrival date; + - An interline ticket (connecting ticket) with the confirmed departure date and seat number leaving for a third country or region within 240 hours; + + - **Visa-free entry ports applicable for this event:** + + - Personnel meeting the above conditions may enter visa-free at either of the two ports: Beijing Capital International Airport or Beijing Daxing International Airport, and must exit from a port that applies the 240-hour visa-free transit policy; + - 🌟 Foreigners transiting to China under the visa-free policy may travel across provinces within the permitted stay and activity areas in 24 provinces (autonomous regions, municipalities). + - Nationals of 38 countries including Brunei, France, Germany, Italy, Spain, Holland, Malaysia, Switzerland, Ireland, Hungary, Austria, Belgium, Luxembourg, New Zealand, Australia, Poland, Portugal, Greece, Cyprus, Slovenia, Slovakia, Norway, Finland, Denmark, Iceland, Andorra, Monaco, Liechtenstein, Korea, Bulgaria, Romania, Croatia, Montenegro, North Macedonia, Malta, Estonia, Latvia, Japan holding valid ordinary passports can be exempted from visa requirement if entering into China for the purpose of business, tourism, family or friends visits, exchange and transit. They can stay in China for no more than 30 days without visa. + +**!! NOTE: The information provided above is for reference only. For specific visa policies applicable to you, please check out** [the website of Chinese Visa Application Service Center](https://www.visaforchina.cn/globle/). + +If, after understanding all the information above, you believe you still need an invitation letter, please send the following information to Willem Jiang, at [ningjiang@apache.org](mailto:ningjiang@apache.org). If all of your information is present, you will receive an invitation letter within 10 business days. + +### INFORMATION NEEDED +- Complete name +- Passport ID +- Gender +- Nationality +- Date of birth +- The name of the Chinese embassy or consulate where to apply for a visa. +- The company represented and the person’s position +- The person’s address and telephone number +- The date the person intends to visit China, include the arraval and departure time +- Who supports the person’s travel experience \ No newline at end of file diff --git a/content/guide/visa_letter_request.zh.md b/content/guide/visa_letter_request.zh.md new file mode 100644 index 000000000..e864fc800 --- /dev/null +++ b/content/guide/visa_letter_request.zh.md @@ -0,0 +1,43 @@ +--- +title: 申请签证 +weight: 20 +--- + +如您需签证邀请函，请参照本页的指示操作。 + +### 说明 + +在申请签证邀请函之前，请先访问您所在城市的[中国签证申请服务中心网站](https://www.visaforchina.cn/globle/)了解相关签证申请信息。另请关注中国最新的签证政策： + +### 请关注中国的最新签证政策： + + - **外籍公民 240 小时免签证过境政策：允许部分国家的公民在不事先申请签证的情况下，享受 10 天的时间前往中国的特定地区。** + + - 适用于下列 54 个国家公民：德国、奥地利、比利时、捷克、丹麦、爱沙尼亚、芬兰、法国、希腊、匈牙利、冰岛、意大利、拉脱维亚、立陶宛、卢森堡、马耳他、荷兰、波兰、葡萄牙、斯洛伐克、斯洛文尼亚、西班牙、瑞典、瑞士、俄罗斯、英国、爱尔兰、塞浦路斯、保加利亚、罗马尼亚、乌克兰、塞尔维亚、克罗地亚、波黑、黑山、北马其顿、阿尔巴尼亚、摩纳哥、白俄罗斯、挪威、美国、加拿大、巴西、墨西哥、阿根廷、智利、澳大利亚、新西兰、韩国、日本、新加坡、文莱、阿联酋、卡塔尔。 + + - **根据中国国家移民局（NIA）的要求，申请 240 小时免签证过境的人必须具备：** + + - 自抵达之日起三个月以上有效的有效护照或其他国际旅行证件 + - 240 小时内确定日期和座位前往第三国家/地区的联程机票。 + + - **适用本次活动的免签入境口岸：** + + - 符合以上条件的人员可免签从北京首都国际机场口岸、北京大兴国际机场口岸 2 个口岸中的任意一个入境，并从适用 240 小时过境免签政策的口岸出境； + - 🌟 过境免签政策来华的外国人可在 24 个省（区、市）允许停留活动区域内跨省域旅行。 + - 文莱、法国、德国、意大利、西班牙、荷兰、马来西亚、瑞士、爱尔兰、匈牙利、奥地利、比利时、卢森堡、新西兰、澳大利亚、波兰、葡萄牙、希腊、塞浦路斯、斯洛文尼亚、斯洛伐克、挪威、芬兰、丹麦、冰岛、安道尔、摩纳哥、列支敦士登、韩国、保加利亚、罗马尼亚、克罗地亚、黑山、北马其顿、马耳他、爱沙尼亚、拉脱维亚、日本 38 国持普通护照人员来华经商、旅游观光、探亲访友、交流访问、**过境不超过30天，可免签入境**。 + +**!! 注意: 以上信息仅供参考，适用您的具体签证政策，请见** [中国签证申请服务中心](https://www.visaforchina.cn/globle/). + +如您了解了以上全部信息，仍确认需要邀请函，请将以下信息发送至 Willem Jiang，邮箱地址为 [ningjiang@apache.org](mailto:ningjiang@apache.org). 如果您的信息齐全无误，我们将在 10 个工作日内向您发送邀请函。 + +### 所需信息 +- 您的完整姓名 +- 您的护照编号 +- 您的性别 +- 您的国籍 +- 您的出生日期 +- 您计划申请签证的中国大使馆或领事馆名称 +- 您所代表的公司及职位 +- 您的个人住址和电话号码 +- 您拟访问中国的日期，包含到达以及离开时间 +- 您的差旅由谁负责 \ No newline at end of file diff --git a/content/search/search.md b/content/search/search.md new file mode 100644 index 000000000..e69de29bb diff --git a/content/search/search.zh.md b/content/search/search.zh.md new file mode 100644 index 000000000..e69de29bb diff --git a/content/sessions/1000.md b/content/sessions/1000.md deleted file mode 100755 index d381ac0cf..000000000 --- a/content/sessions/1000.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Geospatial Indexing and Search at Scale with Apache Lucene" -date: "2019-09-09T15:00:00" -track: "geospatial" -presenters: "Nick Knize" ---- - -Come have a look under the covers at new data structures that enable geospatial and multi-dimensional indexing and search at massive scale in Apache Lucene. This talk will cover the indexing structures considered and ultimately implemented in the Apache Lucene Open Source Project along with the 25 - 30X boost in performance and centimeter spatial accuracy achieved in the latest release. Join us and see what's next for scalable Geospatial Search in Apache Lucene. \ No newline at end of file diff --git a/content/sessions/1001.md b/content/sessions/1001.md deleted file mode 100755 index 3475f52a2..000000000 --- a/content/sessions/1001.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Let's Encrypt Apache Tomcat" -date: "2019-09-10T12:15:00" -track: "tomcat" -presenters: "Christopher Schultz" ---- - -The Let's Encrypt initiative has changed the landscape for online security. The capital cost of cryptographic security for every web site worldwide has been lowered from potentially thousands of dollars per year to zero. One of the trade-offs for the availability of these free certificates is that they are short-lived, making automated deployment of certificates an essential piece of any deployment plan. Let's Encrypt supports Apache httpd out of the box for single-command renewal and deployment on a single host, while with Tomcat it's more complicated. We'll be discussing how Let's Encrypt handles domain-validated authentication, initial requests, and renewals, and how to automate the deployment process in your Tomcat-based environment with minimal downtime. \ No newline at end of file diff --git a/content/sessions/1002.md b/content/sessions/1002.md deleted file mode 100755 index f62dc9656..000000000 --- a/content/sessions/1002.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Using Apache Traffic Control to cache any web object at scale" -date: "2019-09-10T11:15:00" -track: "contentdelivery" -presenters: "Jeff Elsloo" ---- - -This talk will focus on how any organization can deploy a Content Delivery Network (CDN) to cache web object at scale using Apache Traffic Control (ATC) and Apache Traffic Server (ATS). Topics covered will be a brief introduction to the projects, a high-level overview of traffic routing on a CDN, followed by a deep dive into how to efficiently cache objects on an ATC CDN using ATS. Based on years of experience operating ATC CDNs, several different scenarios with corresponding object types (video, web content, etc) will be presented. Learn how the ATC configuration relates to ATS, how different storage mediums can be employed to cache different objects types, how to select the correct redundancy strategies, and how a caching hierarchy can be employed to increase cache efficiency. \ No newline at end of file diff --git a/content/sessions/1003.md b/content/sessions/1003.md deleted file mode 100755 index 8044374d2..000000000 --- a/content/sessions/1003.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Locking Down Apache Tomcat: Practical Security for Real-world Applications" -date: "2019-09-10T17:00:00" -track: "tomcat" -presenters: "Christopher Schultz" ---- - -Out of the box, Apache Tomcat is quite secure. Then you need to configure it to suit your environment, connect your data sources, and deploy your applications. Those processes can potentially reduce the security of the entire system. A thorough review of your host, network, application and configuration is necessary to identify those areas where your security needs improvement. We’ll discuss each of these areas in some detail and how some simple tweaks and tools can make you and your users safer. \ No newline at end of file diff --git a/content/sessions/1004.md b/content/sessions/1004.md deleted file mode 100755 index 5a7c19736..000000000 --- a/content/sessions/1004.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Packaging Tomcat for Linux Distributions" -date: "2019-09-11T12:00:00" -track: "tomcat" -presenters: "Coty Sutherland" ---- - -What is the difference between the ASF Tomcat distribution and the various flavors of tomcat packages on Linux? Most users don’t know! In this discussion Coty will explain how the tomcat packages distributed by a couple flavors of Linux (Red Hat and Debian/Ubuntu) are structured and why the container is provided that way. Coty also hopes to get feedback from the audience and to discuss how these Linux distributions can better provide tomcat to users. \ No newline at end of file diff --git a/content/sessions/1005.md b/content/sessions/1005.md deleted file mode 100755 index 806d718ec..000000000 --- a/content/sessions/1005.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Watching TV on the command line with Apache Traffic Control" -date: "2019-09-10T18:00:00" -track: "contentdelivery" -presenters: "Mark Torluemke" ---- - -Watch closely as a CDN built with the components of Apache Traffic Control comes alive in the delivery of fragmented ABR video. This session will walk through every step in a CDN transaction using command line tools that simulate what real-world clients do behind-the-scenes. \ No newline at end of file diff --git a/content/sessions/1006.md b/content/sessions/1006.md deleted file mode 100755 index ae1420ff5..000000000 --- a/content/sessions/1006.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "How Netflix debugs and fixes Apache Cassandra when it breaks" -date: "2019-09-12T09:00:00" -track: "cassandra" -presenters: "Joey Lynch" ---- - -Netflix relies on Apache Cassandra as a critical source of truth database, and while Cassandra is a remarkably resilient database, it does, ever so occasionally, break. This talk explores how complex Cassandra deployments fail in production, but more importantly, the techniques, tools, and approaches our distributed systems engineers use to debug and mitigate these failures. We will first cover software-based failure modes that come either from our software or the software that Cassandra builds upon. For example, retry storms or unbounded queues can effectively overwhelm the database. Cassandra and Linux give you many tools to detect these, and there are numerous strategies you can use to avoid them. Along the way we will also visit common JVM failure modes and how to assess and remediate these. Next, we will cover hardware-based failure modes that come from a typical cloud environment. We will cover the broad classes of drive and network failures that we cope with every day, as well as the metrics and tests we use to detect them. With these understandings, we will learn how to automatically heal these failures and safely meet our latency SLOs. Finally, we will cover systemic failure modes, including complex interactions of multiple components and systems. This will involve a number of concrete failures we observed that involved multiple bugs, system failures, or data modeling issues in combination. At the end of the talk, we hope that the audience leaves understanding how large distributed databases can fail, but also how to use DevOps skills and tools to debug, mitigate, and automate away these failure modes going forward. \ No newline at end of file diff --git a/content/sessions/1007.md b/content/sessions/1007.md deleted file mode 100755 index ba65acf2d..000000000 --- a/content/sessions/1007.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Kubernetes on BareMetal with Apache CloudStack" -date: "2019-09-10T17:00:00" -track: "cloudstack" -presenters: "Harikrishna Patnala, Suresh Kumar Anaparti" ---- - -nCloudStack currently supports Kubernetes deployment using Hypervisors, which enables PaaS on top of IaaS. However, PaaS applications having data-heavy loads are the most likely to be negatively impacted, especially when the service is oversubscribed. This is where BareMetal cloud offers a significant advantage over Hypervisors, especially for latency-sensitive workloads. This talk and demo will provide turnkey solution for integrating Kubernetes on CloudStack using BareMetal servers. \ No newline at end of file diff --git a/content/sessions/1008.md b/content/sessions/1008.md deleted file mode 100755 index 10916d5cd..000000000 --- a/content/sessions/1008.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Supporting Cassandra In-House – Our Story !!" -date: "2019-09-11T12:00:00" -track: "cassandra" -presenters: "Laxmikant Upadhyay, Anuj Wadhera" ---- - -In this presentation, senior Cassandra Architects at Ericsson will share their vast experience in supporting around 100 Cassandra deployments in production. They will share the key challenges and best practices with respect to Cassandra operations, maintenance and support. There is plenty to learn when the team talks about many problems which they faced in production and how they fixed each one of them successfully by providing interesting solutions. Audience: All Cassandra users especially Cassandra operators and administrators. \ No newline at end of file diff --git a/content/sessions/1009.md b/content/sessions/1009.md deleted file mode 100755 index 35a104d8b..000000000 --- a/content/sessions/1009.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Customizing virtual machines with cloud init" -date: "2019-09-10T14:30:00" -track: "cloudstack" -presenters: "Marcus Sorensen" ---- - -This presentation will cover the basic design around how CloudStack implements support for cloud-init datasources, configuring the cloud-init service within a VM template to talk to one of the CloudStack supported datasources, and give a basic introduction to using cloud-init to customize virtual machines upon deployment. \ No newline at end of file diff --git a/content/sessions/1010.md b/content/sessions/1010.md deleted file mode 100755 index b1e6cc403..000000000 --- a/content/sessions/1010.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "How Netflix manages petabyte scale Apache Cassandra in the cloud" -date: "2019-09-10T11:15:00" -track: "cassandra" -presenters: "Joey Lynch, Vinay Chella" ---- - -At Netflix, we manage petabytes of data in Apache Cassandra which must be reliably accessible to users in mere milliseconds. To achieve this, we have built sophisticated control planes that turn our persistence layer based on Apache Cassandra into a truly self-driving system. We will start with the user interface that Netflix developers use to interact with their Cassandra databases and dive deep into the automation that powers it all. From cluster creation, through scaling up, to cluster death, complex automation drives large fleets of virtual machines hosted on the AWS cloud. First, we will cover the basics of how Netflix deploys Apache Cassandra. In particular, this begins with how we mold Apache Cassandra to the Netflix philosophy of immutable infrastructure, including managing software and hardware upgrades in the face of ever-failing hardware. Then we will explore the concrete techniques needed for such a massive deployment, specifically pull-based control planes and auto-healing strategies. Next, we will cover how Netflix has automated complex but critical Apache Cassandra maintenance tasks such as continuous snapshot backups and always-on anti-entropy repair for keeping our datasets safe and consistent. Both of these systems have gone through multiple architectural evolutions, and we have learned many lessons along the way. Lastly, we will share some of the ways this has gone wrong, and what you can do to avoid them. We will cover a few case studies of major Cassandra outages at Netflix, their root cause, and what we learned from those incidents. At the end of this talk, we hope that participants leave with concrete understanding of the challenges in running massive scale Apache Cassandra as well as solid advice and techniques for building their own self-driving data persistence layer. \ No newline at end of file diff --git a/content/sessions/1011.md b/content/sessions/1011.md deleted file mode 100755 index 22e522728..000000000 --- a/content/sessions/1011.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Next Generation Cassandra Compaction, Going beyond LCS" -date: "2019-09-10T12:15:00" -track: "cassandra" -presenters: "Joey Lynch" ---- - -Cassandra users with general purpose workloads typically must choose the lesser of two evils between the default “write optimized” Size Tiered Compaction Strategy (STCS) and the “read optimized” Leveled Compaction Strategy (LCS). It is typical these days to hear “use LCS unless it explodes, in which case use STCS…unless that explodes, in which case you are out of luck”. I believe that by taking what we as a community learned from the development and real world deployment of LCS, we can create a next generation compaction strategy which offers all of the benefits of LCS with few of the drawbacks. I will begin this talk with an overview of the existing general purpose options, and explore their deficiencies. In particular, we will explore how STCS naïvely groups SSTables of similar size together regardless of read amplification, expends system resources compacting data that does not benefit from compaction, and creates large files that are difficult to process both by Cassandra and downstream systems. The LCS compaction algorithm was developed to mitigate many of these issues. However, LCS suffers from issues of its own: implementation bugs caused by the complexity of the algorithm, significant write amplification introduced in the L0 to L1 transition, and a lack of efficient full compaction. LCS is a worthwhile compaction strategy, but these flaws prevent it from being applied universally to general purpose workloads. Once we have a firm understanding of the issues facing current implementations, I will present a number of potential incremental improvements to the existing choices. I’ll also synthesize these ideas into a blueprint for a new, general purpose, compaction strategy. A key design consideration is to offer users a single compaction interface with easy to understand tunables that adapt to all use cases historically covered by either STCS or LCS. Once I have laid out the proposal, I will open the topic for feedback from other developers in the room and I hope that at the end we will have built consensus around the key challenges of Cassandra compaction, as well as come to a decision on the best path forward. \ No newline at end of file diff --git a/content/sessions/1012.md b/content/sessions/1012.md deleted file mode 100755 index 9386246db..000000000 --- a/content/sessions/1012.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Managing Hundreds of Petabytes of data in the Cloud" -date: "2019-09-10T12:15:00" -track: "bigdata" -presenters: "Lohit VijayaRenu" ---- - -Twitter's on premise HDFS clusters are scaled out to handle hundreds of petabyte data serving thousands of developers every day. Managing such large scale data introduces several challenges. Organization, discovery, data life cycle, replication, schema management and many more questions needs to be answered at scale. Our data infrastructure is built on different open source and in house projects working together to solve this problem. While we extend our data to cloud, we had to solve similar challenges on cloud environment at Twitter scale. In this talk we lay out our framework, talk about challenges we faced on cloud and solutions we are working towards to support thousands of users dealing with hundreds of petabytes of data on cloud. \ No newline at end of file diff --git a/content/sessions/1014.md b/content/sessions/1014.md deleted file mode 100755 index d9db6aff2..000000000 --- a/content/sessions/1014.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "How to Raise an Erudite Chatbot" -date: "2019-09-12T10:00:00" -track: "machinelearning" -presenters: "Boris Galitsky, Jay Taylor" ---- - -Availability of content and training sets is a major bottleneck for a chatbot development today. Relying on Apache OpenNLP and its sub-project OpenNLP.chatbot, we introduce a number of tools and components to design a chatbot and its training set to be knowledgeable and intelligent. n In this talk we will analyze the reasons it is so hard to find a chatbot demo today for a nontrivial task or to observe an intelligent behavior of a chatbot. It is easy to see how a success in AI can boost the chatbot development on one hand, but it is hard to detect intelligence in those chatbots that are available to the public, on the other hand. n We will present an advanced search engine for chatbots with the focus on linguistic features and discourse-level analysis for dialogue management. We will introduce a tool that builds a dialogue from an arbitrary document to form a training dataset for deep learning chatbots. We will demo a chatbot supporting virtual dialogue, where a user joins a virtual community built on the fly, whose members answer questions in this user's current area of interest. An extended content for this talk is available in the book recently published by the speaker “Developing Enterprise Chatbots”. \ No newline at end of file diff --git a/content/sessions/1015.md b/content/sessions/1015.md deleted file mode 100755 index a754edd4b..000000000 --- a/content/sessions/1015.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Using Apache Pulsar to Provide Real-Time IoT Analytics on the Edge" -date: "2019-09-09T15:00:00" -track: "internetofthings" -presenters: "David Kjerrumgaard" ---- - -he business value of data decreases rapidly after it is created, particularly in use cases such as fraud prevention, cybersecurity, and real-time system monitoring. The high-volume, high-velocity datasets used to feed these use cases often contain valuable, but perishable, insights that must be acted upon immediately. In order to maximize the value of their data enterprises must fundamentally change their approach to processing real-time data to focusing reducing their decision latency on the perishable insights that exist within their real-time data streams. Thereby enabling the organization to act upon them while the window of opportunity is open. Generating timely insights in a high-volume, high-velocity data environment is challenging for a multitude of reasons. As the volume of data increases, so does the amount of time required to transmit it back to the datacenter and process it. Secondly, as the velocity of the data increases, the faster the data and the insights derived from it lose value. In this talk, we will present a solution based on Apache Pulsar Functions that significantly reduces decision latency by using probabilistic algorithms to perform analytic calculations on the edge. \ No newline at end of file diff --git a/content/sessions/1016.md b/content/sessions/1016.md deleted file mode 100755 index 4ad7f33c7..000000000 --- a/content/sessions/1016.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "There's life in the (old) dog yet: Apache Edgent reinvents itself and Edge Analytics" -date: "2019-09-09T17:30:00" -track: "internetofthings" -presenters: "Julian Feinauer" ---- - -Apache Edgent (formerly Quarks) is a Project in the Apache Incubator since early 2016. The main focus of the project is to simplify edge processing pipelines. Through the last year it went a bit silent around the project and although the topic is hotter than ever the project lost its drive and went into hibernation. With the rise of other projects in the (industrial) IoT ecosystem like Apache PLC4X and Apache IoTDB there are a lot of new use cases and applications for edge processing, especially in the field of industrial IoT thus, a group of contributors from other IoT related projects has gathered to bring the project back to life and on track.nIn this talk, an introduction into Apache Edgent is given. Its past, its current state and plans on how the project could evolve in the near future.nProjects like Apache PLC4X (incubating) allow us to gather massive amounts of data from machines and plants and close the gap between the `Shopfloor`and the Big Data ecosystem. But, there are a few important differences between `classical` stream processing and the processing of industrial data streams. As the `classical` Big Data world is all about (vertical) scaling and parallelism this is not necessarily true for industrial data. Furthermore, many industrial sides are connected with low bandwidths which makes it even more difficult to send all the data to the cloud. Apache Edgent tries to address these issues by providing a platform for Edge processing. \ No newline at end of file diff --git a/content/sessions/1017.md b/content/sessions/1017.md deleted file mode 100755 index bf1044556..000000000 --- a/content/sessions/1017.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Anomaly Prediction in Apache CloudStack" -date: "2019-09-10T18:00:00" -track: "cloudstack" -presenters: "Suresh Kumar Anaparti, Suresh Sadhu" ---- - -There is a chance of system, storage and network failures in the CloudStack environment at any point of time, which is unknown to the end users and administrators. The administrators are highly depend on the alert and monitoring system to raise an alarm on failures, upon the action is being taken. This can down the operations and will impact the end user services. The prediction of such failures ahead in CloudStack environment would be required instead of depending on the monitoring systems. Anomaly Prediction requires a model using the existing knowledge base and history of the CloudStack environment. This model would help in predicting the failures/abnormalities ahead of time. The necessary action is taken on these to minimize the down time or no down time, and making the CloudStack environment robust. The actions taken are learnt by the system and would automatically apply on the respective failures, make it a self-healing system. \ No newline at end of file diff --git a/content/sessions/1020.md b/content/sessions/1020.md deleted file mode 100755 index ddfdc3a13..000000000 --- a/content/sessions/1020.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Automating Apache Traffic Control Server Deployments with IPv6" -date: "2019-09-09T10:00:00" -track: "contentdelivery" -presenters: "Jason Tucker" ---- - -Apache Traffic Control has some built-in tools for assisting in the installation and deployment of new servers; however, it is still a somewhat manual and repetitive process. I will be demonstrating how the Traffic Control API, used in conjunction with IPv6 autoconf can be used to streamline the deployment process, allowing servers to configure themselves post-install. Will highlight the use of the ‘tc-netconfig’ application, which was developed to facilitate this process. \ No newline at end of file diff --git a/content/sessions/1021.md b/content/sessions/1021.md deleted file mode 100755 index 218431f91..000000000 --- a/content/sessions/1021.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "The Rise of Open Source Software (and Apache projects) in China" -date: "2019-09-11T14:15:00" -track: "community" -presenters: "Ted Liu" ---- - -This talk will cover the evolution, the current status and the outlook of the China open source landscape and ecosystem. This talk will also share the summary and analysis of varies open source software survey reports for China, the largest software developers market in the world. The Apache Software Foundation's projects and their communities play a critical role in leading and developing the OSS ecosystems in China where some leading IT enterprises and individuals, such as Huawei, Alibaba, Baidu, Tencent, Kyligence, SkyWalking, etc., are leading the way in contributing to many ASF projects, open-sourcing their projects, donating to and incubating their projects at ASF, graduating as top-level project of ASF, developing their respective ASF project communities in China, last but not least, advocating the Apache Way when and wherever possible. This reach-out and engage-in journey, starting 2014, to build and cross the culture, language and diversity bridge does not come easy. A story of sweat and tears to share and interact with the audience. \ No newline at end of file diff --git a/content/sessions/1022.md b/content/sessions/1022.md deleted file mode 100755 index b084d4ed2..000000000 --- a/content/sessions/1022.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Managing Trillions of Rows with Aplomb (well, actually with Drill)" -date: "2019-09-11T11:00:00" -track: "drill" -presenters: "Ted Dunning" ---- - -'Ingesting lots of data isn’t very hard any more. Ingesting it on a critical schedule, within strict time bounds while minimizing the risk of bogus data showing up is much harder. In practice, grownup data ingestion and access requires the following capabilities * Incoming data can be fully ingested into our working dataset but hidden from users until all quality checks are completedn * Individual batches of data can be released atomicallyn * Any indexing updates should also appear appear atomicallyn * Expiring data should disappear atomically either according to ingest batch or precise time bounds Apache Drill provides several capabilities that make it much easier to meet these goals. You can handle large volumes of data while allowing in-situ quality controls and while controlling the visibility of unverified data. I will describe a worked example that shows how Drill helps make this happen. (with aplomb)n' \ No newline at end of file diff --git a/content/sessions/1023.md b/content/sessions/1023.md deleted file mode 100755 index ee68af54e..000000000 --- a/content/sessions/1023.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "The State of Diversity in Open Source & the ASF" -date: "2019-09-11T11:00:00" -track: "community" -presenters: "Griselda Cuevas Zambrano" ---- - -I'd like to cover the landscape of D&I in tech, then in open source, and then in the ASF. I'd like to discuss the dimensions used to evaluate our current status and I'd like to discuss the efforts we're working on to changer things. I will like to give this talk in the general track to a) celebrate the 20th Anniversary by reflecting back in our wins on diversity and inclusion and b) make it part of the foundation's big picture \ No newline at end of file diff --git a/content/sessions/1024.md b/content/sessions/1024.md deleted file mode 100755 index 4140e4292..000000000 --- a/content/sessions/1024.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Lessons from the Apache Beam Journey" -date: "2019-09-11T15:15:00" -track: "community" -presenters: "Kenneth Knowles" ---- - -Apache Beam is a big data processing framework that can run on multiple big data engines, including Apache Spark, Apache Flink, Apache Samza, Apache Apex, and a number of non-Apache OSS engines and proprietary engines. After situating Beam in the OSS ecosystem, I will share lessons from Apache Beam's journey, including topics such as: - how do you onboard newcomers to the Apache Way?n - how is open source different / the same as internal corporate work?n - what are techniques for mitigating corporate politics in OSS governance?n - what does a healthy project look like in the stats?n - building community via pull requestn - how do OSS contributing teams look different?n - technical decisions have community impactn - how does ASF merit sustain projects for the long term?n - what seems to motivate volunteers?n - honoring non-code contributions \ No newline at end of file diff --git a/content/sessions/1025.md b/content/sessions/1025.md deleted file mode 100755 index 63c84249c..000000000 --- a/content/sessions/1025.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Declarative Benchmarking of Cassandra and Its Data Models" -date: "2019-09-11T14:15:00" -track: "cassandra" -presenters: "Monal Daxini" ---- - -You have made changes to Cassandra code base. How do you benchmark these changes for scalability and correctness, including different data models (schema), easily? You have created a Cassandra schema for your service. How do you ensure this is scalable? How can you emulate application specific CQL queries, with specified distribution, to validate scale of your schema and associated data scalability without having to code your whole application? I am the author of the NDBench CQL Plugin tool, which was built at Netflix to address these needs and more, Declaratively. One of the Cassandra committers has called this tool 'Game Changing'. This plugin is currently (June 2019) being prepped to be open sourced ahead of this talk. This talk presents: 1. An example of how we used declarative benchmarking to achieve 1 Million requests per second for a user specified data model and query distribution backing a critical service at Netflix. 2. For users: n a. How to certify scalability of new or existing data models on new version of Cassandra, for confident upgrades?n b. How to certify scalability of new complex data models? 3. For committers: n a. Define various profiles to emulate real world use cases to build confidence in changes and certifying new releases. n b. Compare scalability of the same data model across different Cassandra version. 4. The philosophy of the tool, it’s internal architecture, and future enhancements. \ No newline at end of file diff --git a/content/sessions/1026.md b/content/sessions/1026.md deleted file mode 100755 index 434897886..000000000 --- a/content/sessions/1026.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Instaclustr's Open Source Tools For Cassandra - LDAP/Kerberos, Prometheus Exporter, Debug Tooling and K8s Operator" -date: "2019-09-11T16:45:00" -track: "cassandra" -presenters: "Adam Zegelin" ---- - -This session walks devs through Instaclustr's Cassandra tools and how they add key functionality and ease-of-use to their deployments.n- An LDAP authenticator plug-in for CassandranThe open source LDAP authenticator plug-in works closely with the existing CassandraAuthorizer implementation. The plug-in enables developers to quickly reap the benefits of secure LDAP authentication without the need to write their own solutions, and to transition to using the authenticator with zero downtime. - A Kerberos authenticator plug-in for CassandranThe open source Kerberos authenticator plug-in enables Cassandra users to leverage Kerberos’ industry-leading secure authentication and true single sign-on capabilities. The open source project also includes a Kerberos authenticator plugin for the Cassandra Java driver. - Cassandra Prometheus ExporternThe cassandra-exporter is a high-performance metrics collection agent that allows for easy integration with the Prometheus monitoring solution. It has been designed to collect detailed metrics on production-sized clusters with complex schemata with minimal performance impact while at the same time following Prometheus's best practices for exporting metrics. - Additional utilities and debugging toolsnIncluding a utility to assist with backup and restore to various cloud providers, and tools to provide debug-level information about SSTables. - A Cassandra operator for running and operating Cassandra within KubernetesnThe open source Cassandra operator functions as a Cassandra-as-a-Service on Kubernetes, fully handling deployment and operations duties so that developers don’t have to. It also offers a consistent environment and set of operations founded on best practices, which is reproducible across production clusters and development, staging, and QA environments. The audience for this presentation will learn the specifics of how to implement – and get the most out of – these open source solutions. \ No newline at end of file diff --git a/content/sessions/1027.md b/content/sessions/1027.md deleted file mode 100755 index 2459ccaf3..000000000 --- a/content/sessions/1027.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Two years in the making: What’s new with Apache Cassandra 4.0?" -date: "2019-09-11T11:00:00" -track: "cassandra" -presenters: "Dinesh Joshi" ---- - -At almost two years in the making Apache Cassandra 4.0 is almost here. With a focus on performance and stability, it is full of interesting features. This talk takes you through a tour of the new features and performance improvements. From Zero Copy Faster Streaming support to Virtual Tables and Audit Logging, learn how Apache Cassandra 4.0 has departed from previous major releases to be focused on the features required by the most demanding users. Attendees will not only gain an understanding of new features in Apache Cassandra, but understand what drove their creation and therefore have a better idea of how to best use these features in their own systems. \ No newline at end of file diff --git a/content/sessions/1028.md b/content/sessions/1028.md deleted file mode 100755 index c60075b77..000000000 --- a/content/sessions/1028.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Happiness is a hybrid cloud with Apache Cassandra™" -date: "2019-09-11T17:45:00" -track: "cassandra" -presenters: "Patrick McFadin" ---- - -The world's data and the speed with which we expect to access it continues to grow rapidly. Apache Cassandra was built for hybrid cloud implementations and that makes it easy for companies to fully exploit their cloud strategies without having to re-architect their applications or change their systems for data management. Sounds too good to be true? A lot of big companies have been doing it for years and you can too! In this presentation, we will cover best practices and strategies for deploying your app to Cassandra on any cloud that you want. We will discuss: * Uses cases that require the power of Cassandra for hybrid cloud * Data model considerations for your active anywhere applicatio * Deployment strategies on any cloud. We do not care which cloud :) \ No newline at end of file diff --git a/content/sessions/1030.md b/content/sessions/1030.md deleted file mode 100755 index 3f3e7b3c8..000000000 --- a/content/sessions/1030.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "How eBay build Data Quality Platform with Apache Griffin" -date: "2019-09-10T14:30:00" -track: "observability" -presenters: "William Guo, Branky Shao" ---- - -High Quality data plays an important role for incubating awesome big data applications. In real projects, governing data, observing their quality health, and taking remedy actions are tedious and time-consuming. Apache Griffin is an open source Data Quality solution for Big Data, which supports both batch and streaming mode. It offers a unified process to measure your data quality from different perspectives, helping you build trusted data assets, therefore boost your confidence for your business. In this session, we will talk about how eBay leverage Apache Griffin and build a big data quality platform to support all kinds of business critical data applications. First, we will introduce the key concepts and architecture design of Griffin and share the experience of open-source it under the Apache way. Then, we will dive into our internal data quality platform and its use cases. \ No newline at end of file diff --git a/content/sessions/1031.md b/content/sessions/1031.md deleted file mode 100755 index 6627bafab..000000000 --- a/content/sessions/1031.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Improving the Observability of Cassandra, Kafka and Kubernetes applications with Prometheus and OpenTracing" -date: "2019-09-10T15:30:00" -track: "observability" -presenters: "Paul Brebner" ---- - -As distributed applications grow more complex, dynamic, and massively scalable, “observability” becomes more critical. Observability is the practice of using metrics, monitoring and distributed tracing to understand how a system works. In this presentation we’ll explore two complementary Open Source technologies: Prometheus for monitoring application metrics; and OpenTracing and Jaeger for distributed tracing. We’ll discover how they improve the observability of a massively scalable Anomaly Detection system - an application which is built around Apache Cassandra and Apache Kafka for the data layers, and dynamically deployed and scaled on Kubernetes, a container orchestration technology. We will give an overview of Prometheus and OpenTracing/Jaeger, explain how the application is instrumented, and describe how Prometheus and OpenTracing are deployed and configured in a production environment running Kubernetes, to dynamically monitor the application at scale. We conclude by exploring the benefits of monitoring and tracing technologies for understanding, debugging and tuning complex dynamic distributed systems built on Kafka, Cassandra and Kubernetes. \ No newline at end of file diff --git a/content/sessions/1032.md b/content/sessions/1032.md deleted file mode 100755 index c31706de6..000000000 --- a/content/sessions/1032.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Panel discussion on recognizing non-code contributions" -date: "2019-09-10T14:30:00" -track: "community" -presenters: "Griselda Cuevas Zambrano" ---- - -A discussion panel with 2 or 3 projects who have been successful at growing their communities outside the English speaking countries and how they did it. \ No newline at end of file diff --git a/content/sessions/1033.md b/content/sessions/1033.md deleted file mode 100755 index 3527b4224..000000000 --- a/content/sessions/1033.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache Groovy Update and Roadmap" -date: "2019-09-12T13:00:00" -track: "general" -presenters: "Paul King" ---- - -This talk looks at the latest features in Groovy from 2.5 and 3.0. This includes new AST transforms, new macro features, the new Parrot parser and a myriad of other new miscellaneous features. The talk outlines a broad roadmap of how the new features are planned to be rolled out and the system requirements for each version. \ No newline at end of file diff --git a/content/sessions/1034.md b/content/sessions/1034.md deleted file mode 100755 index f0ec59e3c..000000000 --- a/content/sessions/1034.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Efficient data integrations with the Medley platform" -date: "2019-09-11T16:45:00" -track: "integration" -presenters: "Sami Adranly" ---- - -In this talk, we will introduce the Medley data integration platform used at Uber. Medley is based on the Apache Camel framework, and centered around integrating internal systems efficiently and with great precision on multiple network environments. We will focus on the concept of reusability of integrations, where integrations can be securely shared by implementing integrations as web services. Also, implementing business entity change detection with publish/subscribe to changes to allow event based integrations based on data state changes. Finally, we will describe the challenges dealing with many types of integration use cases, development processes, monitoring, and troubleshooting. \ No newline at end of file diff --git a/content/sessions/1035.md b/content/sessions/1035.md deleted file mode 100755 index b75b818ca..000000000 --- a/content/sessions/1035.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "State of the cat, new and upcoming" -date: "2019-09-10T11:15:00" -track: "tomcat" -presenters: "Remy Maucherat" ---- - -This presentation will detail the current state of the Apache Tomcat project and its community. After going through information on recent releases, community metrics and security advisories, this will cover important new and upcoming features. This includes HTTP/X support, clustering, TLS, cloud enablement, and much more. \ No newline at end of file diff --git a/content/sessions/1036.md b/content/sessions/1036.md deleted file mode 100755 index acf0356c7..000000000 --- a/content/sessions/1036.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Introducing Secure Boot in Apache CloudStack" -date: "2019-09-10T18:00:00" -track: "cloudstack" -presenters: "Pavan Kumar Aravapalli, Hari Krishna Patnala" ---- - -nExisting CloudStack supports legacy boot loader BIOS by default for Guest VM’s booting on Hypervisor(XEN, VMware, KVM etc). In Virtualization, hypervisor emulates the hardware for the Guest VM. BIOS {Basic Input-Output system} is a low-level software. The BIOS loads when a VM starts up, and the BIOS is responsible for waking up your VM’s hardware(emulated) components, ensures they are functioning properly, and then runs the boot loader that boots Windows or whatever other operating system have installed. The existing BIOS has serious limitations. It can only boot from drives of two 1 TB or less. 3 TB drives are now common, and a computer with a BIOS can’t boot from them. That limitation is due to the way the BIOS’s Master Boot Record system works. BIOS can be used in devices that do not require large storage or security. UEFI { Unified Extensible Firmware Interface} can work alongside BIOS. It can sit on top of BIOS and work independently. Benefits of UEFI boot mode over Legacy BIOS boot mode include: Support for hard drive partitions larger than 2 TBn Support for more than four partitions on a driven Fast bootingn Efficient power and system managementn Robust reliability and fault management \ No newline at end of file diff --git a/content/sessions/1037.md b/content/sessions/1037.md deleted file mode 100755 index 9f439cde8..000000000 --- a/content/sessions/1037.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Network Visibility in Hybrid and Multi Clouds" -date: "2019-09-09T17:30:00" -track: "cloudstack" -presenters: "Suresh Kumar Anaparti, Harikrishna Patnala" ---- - -Network Visibility in Apache CloudStack \ No newline at end of file diff --git a/content/sessions/1038.md b/content/sessions/1038.md deleted file mode 100755 index 6ce4a6bcc..000000000 --- a/content/sessions/1038.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Podling's Shark Tank" -date: "2019-09-11T17:45:00" -track: "community" -presenters: "Roman Shaposhnik" ---- - -Is it a panel? Is it a talk? It is a Podling's Shark Tank! Back by popular demand with even sharkier judges! What is it, you ask? Well, this is just like Shark Tank TV show (think speed dating between entrepreneurs and investors) but instead of “Squirrel Boss” and “Man Candle” you'll be hearing pitches for Apache Incubator projects. Also instead of Mark Cuban and Kevin O'Leary you'll be pitching to the panel of ASF old timers (trying to convince them that your project is worthy of their esteemed attention and endorsement). There will be snark, there will be prizes, there will be reciting of Apache Way creed. But most of all there will be fun. We guaranteed it! And don't forget, if you want your favorite podling to be featured on the show, volunteer to do the pitch by simply signing up on this wiki page: https://wiki.apache.org/apachecon/ACNA19PodlingSharkTank \ No newline at end of file diff --git a/content/sessions/1039.md b/content/sessions/1039.md deleted file mode 100755 index 8311d0cfc..000000000 --- a/content/sessions/1039.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Cloudstack API development 101" -date: "2019-09-09T16:30:00" -track: "cloudstack" -presenters: "Marcus Sorensen" ---- - -This is a primer to getting started with adding functionality to CloudStack. In this presentation we will go over a custom Maven archetype that will fast track CloudStack development, whether the goal is to develop functionality to contribute to CloudStack project or to integrate CloudStack with site specific systems. It is geared toward Java developers who are new to the CloudStack code base. \ No newline at end of file diff --git a/content/sessions/1040.md b/content/sessions/1040.md deleted file mode 100755 index c26054e32..000000000 --- a/content/sessions/1040.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Open vSwitch with DPDK on CloudStack" -date: "2019-09-09T14:00:00" -track: "cloudstack" -presenters: "Nicolas Vazquez" ---- - -Support for Open vSwitch with DPDK has been introduced on CloudStack version 4.12. In this talk we will go through the motivations and advantages of using DPDK on KVM hosts, high level architecture of the feature, what is needed and how can it be used on production environments and future works. \ No newline at end of file diff --git a/content/sessions/1041.md b/content/sessions/1041.md deleted file mode 100755 index ce7e34e1b..000000000 --- a/content/sessions/1041.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "The Apache Way May Save Us All" -date: "2019-09-12T13:00:00" -track: "community" -presenters: "Matt Ryan" ---- - -The commercial software industry is plagued by numerous significant problems: Cost and schedule overruns, poor software quality, unrealistic expectations, products that fail to meet customer needs, and more. These problems are historic, chronic, and pervasive; they’ve been with us for decades, and changes to methodology haven’t had much effect in remedying them. By contrast, the Apache Software Foundation has been around for 20 years and the Apache Way - the framework of values and governance for Apache Software Foundation projects - has been the guiding force for dozens of highly successful software projects that are used in countless environments everywhere, run much of the Internet, and are probably used by every technical person in one way or another every day. In fact it would be hard to find any other software organization, open or proprietary, with a better track record. What if the Apache Way were applied to all software projects? Could the Apache Way be the answer to solve these universal software engineering challenges, for everybody, once and for all? \ No newline at end of file diff --git a/content/sessions/1042.md b/content/sessions/1042.md deleted file mode 100755 index 8b2ee5ba3..000000000 --- a/content/sessions/1042.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Cassandra Traffic Management at Instagram" -date: "2019-09-12T17:30:00" -track: "cassandra" -presenters: "Michaël Figuière" ---- - -Cassandra has been deployed for many years at Instagram and is still growing fast. Over the years, we’ve constantly improved the design of our infrastructure: we’ve introduced Cassandra proxy nodes to decouple the processing and storage workloads, we've developed Rocksandra, a new storage engine relying on RocksDB that reduced the GC pressure and improved the efficiency of our clusters. Lately, we've added an intermediate layer in our Cassandra infrastructure, where we could add many traffic optimizations to further improve our Cassandra clusters' efficiency and reliability. This talk will explain in detail this new component and the results we've observed. \ No newline at end of file diff --git a/content/sessions/1043.md b/content/sessions/1043.md deleted file mode 100755 index 68aa15eee..000000000 --- a/content/sessions/1043.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache Training, How it came to be and what it is about" -date: "2019-09-12T15:30:00" -track: "community" -presenters: "Soenke Liebau, Lars Francke" ---- - -There are lots of companies offering training for, and around, the Apache ecosystem, as well as many other topics that all create their own training material. Keeping this material up to date, especially in the fast-moving open-source world is not an easy task and takes a lot of time and effort. And all the time someone not too far away is probably working on a very similar slide-deck to explain what ZooKeeper does... It is this fundamental issue that the Apache Training project was created to address - centralising training resources and making them easier to access and (re-)use. We believe that training material, while important, is only part of what constitutes good training. The larger part is an experienced trainer who actually understands how the key aspects can be applied in real life and also the consequences of not implementing things correctly.nGreat training tends to come from the experiences and stories that a trainer can share in addition to the actual material. Based on this underlying belief and the very real pain of updating training material the Apache Training project was born. In early 2019, it entered Apache Incubator and is already building a community and gaining various contributions. In this talk, we want to share with you some of the events and the thought process that went into the creation of this project, as well as our main goals and principles. The first part of this talk will focus on how we arrived at the decision that open-source training material is a good thing to have and a little bit of the early history of this project. In the second part, we will discuss what the project is actually about. We will demonstrate the version-control friendly system of designing slides in source control that is at the core of this project. On top of these source files, we aim to create a metadata repository which will enable everybody to efficiently search available content and easily create tailored decks from existing material. And last but very much not least, we will cover how you can contribute to and benefit from this project, as this is front and center an effort from the community for the community! \ No newline at end of file diff --git a/content/sessions/1044.md b/content/sessions/1044.md deleted file mode 100755 index 4eb0ea5e6..000000000 --- a/content/sessions/1044.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache Cordova In Action" -date: "2019-09-12T09:00:00" -track: "mobile" -presenters: "Hazem Saleh" ---- - -Apache Cordova is a platform for building Hybrid mobile apps using common Web technologies (HTML, CSS and JavaScript). Apache Cordova offers a set of APIs that allow the mobile app developers to utilize mobile native functions such as (Audio, Camera, Contacts …etc) using JavaScript. Ionic is one of the best mobile web application frameworks, which allows the web developers to develop neat mobile web applications. This session discusses why there is a need for Hybrid mobile development, the current challenges of mobile development, and how using Apache Cordova can help in overcoming many of these technical challenges. Finally, it demonstrates a real Cordova mobile app for showing the audience the best practices of designing, developing, and deploying hybrid Android and iOS mobile apps. \ No newline at end of file diff --git a/content/sessions/1045.md b/content/sessions/1045.md deleted file mode 100755 index ad7cf41b6..000000000 --- a/content/sessions/1045.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Welcome to the Content Delivery Track" -date: "2019-09-09T08:00:00" -track: "contentdelivery" -presenters: "Dave Neuman" ---- - -This presentation will open the track and serve as a welcome to those attending. I will cover what’s new in the ATC (Apache Traffic Control) community, introduce our new committers and PMC members, talk about what we will cover during the conference and then talk a little about the CDN market as a whole. \ No newline at end of file diff --git a/content/sessions/1046.md b/content/sessions/1046.md deleted file mode 100755 index 4d4c8d133..000000000 --- a/content/sessions/1046.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Workshop: Getting Apache Traffic Control Up And Running" -date: "2019-09-09T17:30:00" -track: "contentdelivery" -presenters: "Dan Kirkwood" ---- - -CDN-in-a-Box is a subproject within Apache Traffic Control (ATC) to build a test/experimental CDN in docker on an individual workstation. The process is still challenging, so I'd like to do a workshop to guide interested users through the process on their own laptops. The goal is to show them how the CDN works at a fairly low level and how to test that the various parts work as expected as well as troubleshooting if things go awry. This would hopefully drum up more interest in participation in the ATC project. \ No newline at end of file diff --git a/content/sessions/1047.md b/content/sessions/1047.md deleted file mode 100755 index 9a11736d2..000000000 --- a/content/sessions/1047.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Anycast ATC Integration at Comcast" -date: "2019-09-09T15:00:00" -track: "contentdelivery" -presenters: "Jamie Panagos" ---- - -Anycast is an industry proven technology and is quickly becoming a requirement for off-the-shelf CDNs as well as CDN-as-a-Service. In this talk we will investigate how Comcast is integrating anycast functionality into Apache Traffic Control to improve resiliency, convergence and localization. \ No newline at end of file diff --git a/content/sessions/1048.md b/content/sessions/1048.md deleted file mode 100755 index 223e4f15e..000000000 --- a/content/sessions/1048.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache Tomcat, your webapp, and the Graal" -date: "2019-09-11T11:00:00" -track: "tomcat" -presenters: "Remy Maucherat" ---- - -Graal's native image generation has been the focus of a lot of attention recently, despite being in the very early stages. This session will look at the Tomcat embedded packaging, how to handle your webapp, and will demo use of Graal and Tomcat in a container environment. The session will mostly be tutorial/demo oriented, after a brief overview of the concept and tools used. \ No newline at end of file diff --git a/content/sessions/1049.md b/content/sessions/1049.md deleted file mode 100755 index 2c3564c3a..000000000 --- a/content/sessions/1049.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache Hadoop 3.x State of The Union and Upgrade Guidance" -date: "2019-09-11T11:00:00" -track: "bigdata" -presenters: "Jitendra Pandey, Suma Shivaprasad" ---- - -'Apache Hadoop YARN is the modern Distributed Operating System for big data applications. It morphed the Hadoop compute layer to be a common resource-management platform that can host a wide variety of applications. Many organizations leverage YARN in building their applications on top of Hadoop without themselves repeatedly worrying about resource management, isolation, multi-tenancy issues etc. The Hadoop Distributed File System (HDFS) is the primary data storage system used by Hadoop applications. It employs a NameNode and DataNode architecture to implement a distributed file system that provides high-performance access to data across highly scalable Hadoop clusters. In this talk, we’ll start with the current status of Apache Hadoop 3.x – how it is used today in deployments large and small. We’ll then move on to the exciting present & future of Hadoop 3.x – features that are further strengthening Hadoop as the primary resource-management platform as well as the storage system for enterprise data-centers. We’ll discuss the current status as well as the future promise of features and initiatives for both YARN and HDFS of Hadoop 3.x: For YARN 3.x, we have powerful container placement, global scheduling, support for machine learning (Spark) and deep learning (TensorFlow) workloads through GPU & FPGA scheduling and isolation support, extreme scale with YARN federation, containerized apps on YARN, support for long-running services (alongside applications) natively without any changes, seamless application/services upgrades, powerful scheduling features like application priorities, intra-queue preemption across applications and operational enhancements including insights through Timeline Service V2, a new web UI, better queue management, etc.nAlso, HDFS 3.0 announced GA for erasure coding which doubles the storage efficiency of data and thus reduces the cost of storage for enterprise use cases. HDFS added support for multiple standby NameNodes for better availability. For better reliability of metadata and easier operations, Journal nodes have been enhanced to sync the edit log segments to protect against rolling failures.nDisk balancing within a DataNode was another important feature added to ensure disks are evenly utilized in a DataNode, which also ensures better aggregate throughput, and prevents from lopsided utilization if new disks are added or replaced in a DataNode. HDFS team is currently driving the Ozone initiative which lays the foundation of the next generation of storage architecture for HDFS where data blocks are organized in Storage Containers for higher scale and handling of small objects in HDFS. Ozone project also includes an object store implementation to support new use cases. At last, since more and more users are planning to upgrade from 2.x to 3.x to get all the benefits mentioned above, we will also briefly talk about upgrade guidance from Hadoop 2.x to 3.×.n' \ No newline at end of file diff --git a/content/sessions/1050.md b/content/sessions/1050.md deleted file mode 100755 index 6c660b7b0..000000000 --- a/content/sessions/1050.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Drilling Security Data" -date: "2019-09-11T14:15:00" -track: "drill" -presenters: "Charles Givre" ---- - -Security data is often challenging to analyze because it comes in a variety of formats that are difficult and time consuming to parse, requiring a myriad of various tools to analyze this data. Additionally, bringing this data together to correlate multiple data sets can be difficult and extremely time consuming. But what if all your data could be queried with single tool, using a common language? This talk will demonstrate how to use Apache Drill's enormous analytic power on security data sets and visualize this data using Apache Superset. \ No newline at end of file diff --git a/content/sessions/1051.md b/content/sessions/1051.md deleted file mode 100755 index 5e81ec6ab..000000000 --- a/content/sessions/1051.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Building a smart query mediation engine for Genome Research using Apache Drill" -date: "2019-09-09T11:30:00" -track: "drill" -presenters: "Tushar Pathare" ---- - -The integration of both clinical and genomics data into a single platformnthat will allow the researchers to search, analyze and discover the hidde patterns in biomedical databases. Developing this integrated platform be-ncomes a complex and challenging task due to (i) heterogeneous data typesn(ii) huge volume of data (iii) number of attributes are di erent across thendata (iv) generated data are from heterogeneous platform and (v) corre-nlation between clinical and genomics data are di erent to each other andn(vi) scalability factors. To address these problems, Biomedical InformaticsnDivision within Sidra Research took the initiative to develop a personalized health-nintegrated informatics platform that will help the researchersnto get insights to questions previously un-answered. \ No newline at end of file diff --git a/content/sessions/1052.md b/content/sessions/1052.md deleted file mode 100755 index 4722212e7..000000000 --- a/content/sessions/1052.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "What’s Surprising about Apache Drill and Why That’s a Challenge" -date: "2019-09-11T12:00:00" -track: "drill" -presenters: "Ellen Friedman" ---- - -'Apache Drill has some very surprising characteristics and, more importantly, it enables Drill users to do some surprising things. It’s no longer surprising to be able to do standard SQL in a highly distributed and large scale system - there is an entire class of modern tools that do this including Apache Hive, Presto or Spark SQL. But Drill has other capabilities that are surprising and make it stand apart from its class. For one thing, Drill provides an extraordinary degree of flexibility in several ways, including: Support for a wide variety of file formats including semi-structured and nested data (such as Parquet, JSON, Avro) and non-file data sources - not just data access but ability to fully use these data sources with high performance Schema discovery - a capability that opens up data exploration in unexpected ways for Drill users and allows progressive data modelling Easy extensibility with high performance - you don’t have to trade one for the other These are valuable if surprising capabilities. Why, then, is that a challenge? Because people don’t expect them, they also may not come looking for a tool that can do these things. The challenge comes in how to make potential users aware of the opportunities that Drill offers. This talk will explore some of Drill’s surprising capabilities, how it’s able to do these things and what impact that has for Drill users. In addition, we will open a discussion about how best to inform and engage a broader user community. This latter issue is not only important for Drill but for other Apache projects as well.' \ No newline at end of file diff --git a/content/sessions/1053.md b/content/sessions/1053.md deleted file mode 100755 index 013a439e7..000000000 --- a/content/sessions/1053.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Monitoring of CloudStack" -date: "2019-09-10T14:30:00" -track: "cloudstack" -presenters: "Alexander Stock" ---- - -Our approach of monitoring the CloudStack management server, parts of CloudStacks ecosystem and its underlying components. We give an introduction of how to monitor virtual routers, customer systems and customer resources in CloudStack. \ No newline at end of file diff --git a/content/sessions/1054.md b/content/sessions/1054.md deleted file mode 100755 index 7855138c1..000000000 --- a/content/sessions/1054.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache Cassandra Sidecar, let’s make C* attractive and easy to operate" -date: "2019-09-12T15:30:00" -track: "cassandra" -presenters: "Vinay Chella, Joey Lynch" ---- - -Cloud database offerings have expanded over the past decade, encompassing everything from virtualized machines in the cloud to entirely serverless databases. With this pace of innovation in the cloud ecosystem, Cassandra stands in a unique position to serve its users with unique advantages over any other system. This also puts Cassandra in an interesting position to compete with the ongoing innovations in the cloud. With the internal architecture and storage mechanism aside, citizen developers in the community are looking for several other operability aspects/ ecosystem around the services for the long term investments and benefits from such service. At this juncture, it is vital for the Cassandra dev community to build muscle around supporting such an ecosystem that the community is looking for, to be on par with rest of cloud services and other competing offers in the industry. Specifically, it is important for us to innovate and focus on improving these areas along with the rest of the product. - Ease of use n- Simplified operability in the cloud, precisely simplified operability in the hybrid/multi-cloudn- Pluggability with other infrastructure services such as metrics, discovery, and monitoringn- Painless rollouts of version and protocol upgradesn- Elegant developer experience in polyglot environmentsn- Dev education of Cassandra best practicesn- Unified access across the complex systems As Cassandra stands today, operating the database requires either considerable labor, complex automation, or both. Some of this complexity is an unavoidable result of operating a distributed system, but much of it is operational complexity stemming from properties of C* itself. As a result of these complexities, C* operators spend too much time dealing with issues that the database should solve on its own, and are unable to reap the full benefit of Cassandra’s powerful distributed data model. As part of this talk, we will focus on how to address the aforementioned challenges to keep the Cassandra competitive within the cloud offerings and database services industry, with simplified operability and elegant developer and operator experience. We also hope to get Apache Cassandra 4.0 up and running in any cloud without a hassle with the help of the sidecar. We hope to leave a thought in the Cassandra dev community to start thinking about these areas in upcoming releases of Cassandra. \ No newline at end of file diff --git a/content/sessions/1055.md b/content/sessions/1055.md deleted file mode 100755 index 654d98da1..000000000 --- a/content/sessions/1055.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Agile Integration - Cloud Native Application Development" -date: "2019-09-12T14:00:00" -track: "integration" -presenters: "Christina Lin" ---- - -Cloud native application development is more than just container orchestration, when it comes to designing proper agile software architecture, there are many aspect that need to be taken into account. From simple microservices runtime, orchestration of core business, interacting with legacy, connecting with external SaaS application. To a more reactive system with events driven backbone, and also avoid data silos and how to deal with routing, versioning deployment strategy. Putting everything into a big picture, guide you through what next generation of cloud native architecture should be like and how everything works together. \ No newline at end of file diff --git a/content/sessions/1056.md b/content/sessions/1056.md deleted file mode 100755 index 73d8ce8c5..000000000 --- a/content/sessions/1056.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "The Apache Way - Not only the What but the Why" -date: "2019-09-10T11:15:00" -track: "community" -presenters: "Jim Jagielski" ---- - -The Apache Way consists of the basic core tenets, embraced by all ASF projects, which define how Open Source projects should be managed, governed and maintained. But in addition to a simply list of what those tenets are, we will dig deeper and discover the history and rationale behind them all. The incredible success of the Apache Way can be used to immediate advantage in your open source project and actual forms the core philosophy behind InnerSource. \ No newline at end of file diff --git a/content/sessions/1057.md b/content/sessions/1057.md deleted file mode 100755 index de5ae0619..000000000 --- a/content/sessions/1057.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache httpd 2.4 Overview" -date: "2019-09-12T14:00:00" -track: "general" -presenters: "Jim Jagielski" ---- - -All the Ins and Outs related to what's old, and NEW, in Apache httpd 2.4. If you haven't use httpd in awhile, this is a great way to find out all the new features and enhancements. And if you are a current user, you find out about how to tune httpd for peak performance \ No newline at end of file diff --git a/content/sessions/1058.md b/content/sessions/1058.md deleted file mode 100755 index fe0073350..000000000 --- a/content/sessions/1058.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache SkyWalking -- Top level open source APM" -date: "2019-09-10T11:15:00" -track: "observability" -presenters: "Sheng Wu" ---- - -SkyWalking has just graduated as Top Level Project in this year. SkyWalking is an open source APM system, including monitoring, tracing, diagnosing capabilities for distributed systems in Cloud Native architecture. In the modern world, we have legacy and cloud-native distributed systems, even in hybrid. SkyWalking provides different observability methods, including language agents and service mesh angle, which adopt different scenarios. In this session, I will introduce what SkyWalking do, and how we do it, even without requiring any big data tech. A lightweight and powerful observability platform provides the core features of APM system. \ No newline at end of file diff --git a/content/sessions/1059.md b/content/sessions/1059.md deleted file mode 100755 index fae9d3a8a..000000000 --- a/content/sessions/1059.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache Camel K: connect your Knative serverless applications with everything else" -date: "2019-09-12T09:00:00" -track: "integration" -presenters: "Andrea Tarocchi, Nicola Ferraro" ---- - -When you start developing serverless applications in the real world, sooner rather than later you will need to talk to external (legacy) systems. In this talk you will discover how to leverage Apache Camel K to connect your Knative serverless applications with everything Camel can connect.nApache Camel K allows running Camel routes as serverless applications directly on top of any Kubernetes cluster, leveraging Knative serverless capabilities such as auto-scaling, scaling to zero, event-based communication in order to connect serverless functions and microservice with external systems. \ No newline at end of file diff --git a/content/sessions/1060.md b/content/sessions/1060.md deleted file mode 100755 index c1cefc3cd..000000000 --- a/content/sessions/1060.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "The Apache TinkerPop4 Virtual Machine" -date: "2019-09-11T11:00:00" -track: "graphprocessing" -presenters: "Marko A. Rodriguez" ---- - -The Apache TinkerPop team is working on the next major version of the project -- TinkerPop4. TinkerPop4 advances the technology developed in TinkerPop3 with a new virtual machine architecture that enables any query language to control any processing engine manipulating any data structure. A query language can support a compiler to TinkerPop4's Turing Complete bytecode specification. Compiled bytecode is then translated into an intermediate functional representation that an integrated processing engine will ultimately transform into a processor-specific execution plan. TinkerPop4's intermediate representation naturally supports pull-based, push-based, distributed, message passing, threaded, query routing, etc. semantics which are captured in various data processors such as Pipes, Rx, Akka, Flink, Apex, Spark, and more. Additionally, a subset of the instructions in TinkerPop's bytecode specification denote manipulations to data stored in graph databases, relational databases, document databases, in-memory, and other data structure systems. In this way, TinkerPop4 enables system-agnostic universal computing across the numerous Apache (and other) language, processor, and database projects. This presentation offers an in-depth, technical analysis of the TinkerPop4 virtual machine architecture. \ No newline at end of file diff --git a/content/sessions/1061.md b/content/sessions/1061.md deleted file mode 100755 index bf5f0deef..000000000 --- a/content/sessions/1061.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "GeoSpatial and Temporal Forecasting in Uber Marketplace" -date: "2019-09-09T16:30:00" -track: "geospatial" -presenters: "Chong Sun, Brian Tang" ---- - -Uber’s Marketplace is the algorithmic brain and decision engine behind our ride-sharing services. Marketplace Forecasting builds and deploys ML algorithms to handle the immense coordination, hyperlocal decision making, and learning needed to tackle the enormous scale and movement of our transportation network. In order for our decision engines to be future-aware, we need to be able to “see into the future” as accurately as possible across both space and time. In Uber, we use H3 (a hexagonal hierarchical geospatial indexing system) to partition the data geospatially. To incorporate both geospatial and temporal features into the forecasting models in real time, we need efficient and scalable techniques for data processing. In this presentation, we provide an overview of geospatial and temporal forecasting problem in Uber Marketplace Forecasting. Then, we use real time forecasting as an example to show the need for efficient data smoothing and describe a hexagon convolution based approach for processing H3 data in production. We will demonstrate how we could incorporate hexagon convolution with Apache Flink to scale the data smooth for forecasting. \ No newline at end of file diff --git a/content/sessions/1062.md b/content/sessions/1062.md deleted file mode 100755 index 6f9ca3059..000000000 --- a/content/sessions/1062.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Automating Apache Traffic Control Environment Deployments with Ansible" -date: "2019-09-09T11:30:00" -track: "contentdelivery" -presenters: "Jonathan Gray" ---- - -Apache Traffic Control is a set of applications designed to complement Apache Traffic Server to comprise a Content Delivery Network. Currently the creation of a complete production-ready CDN environment is a complex process. I will be demonstrating automation leveraging Ansible and CDN environment designs to facilitate consistent and more rapid testing environment creation. \ No newline at end of file diff --git a/content/sessions/1063.md b/content/sessions/1063.md deleted file mode 100755 index 9f8d2b5c5..000000000 --- a/content/sessions/1063.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Machine Learning and Data Analytics in the Content Delivery Context" -date: "2019-09-09T16:30:00" -track: "contentdelivery" -presenters: "John Clem" ---- - -In line with the content delivery community's interest in data-driven operational insights and improving operational efficiency, this presentation and discussion will address the following topics: (1) What are Machine Learning, Deep Learning, and the different types of analytics (from descriptive through prescriptive)? (2) What are the potential benefits and risks of analytical solutions? (3) And finally - a walkthrough of some example use cases from modern content delivery networks and their corresponding analytical solutions. \ No newline at end of file diff --git a/content/sessions/1064.md b/content/sessions/1064.md deleted file mode 100755 index 99e0b71ea..000000000 --- a/content/sessions/1064.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Traffic Server Remap Update" -date: "2019-09-10T12:15:00" -track: "contentdelivery" -presenters: "Alan Carroll" ---- - -Rewriting URLs, or 'remapping', is probably the single most used feature of the Traffic Servernproxy. With the transition to YAML from a proprietary line based format, it was decided to startnover with effectively a blank slate, to design a configuration with a functionality emphasizingnpower and flexibility over compatibility. This talk will elaborate on the decision path leading upnto this complete restructuring, the design of the configuration, and examples of its use. Thenunderlying implementation will be sketched out to show some of the interesting C++ techniques usednto make it possible. \ No newline at end of file diff --git a/content/sessions/1065.md b/content/sessions/1065.md deleted file mode 100755 index 8f07ee849..000000000 --- a/content/sessions/1065.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "The ASF and OFBiz - How Community Makes It All Possible" -date: "2019-09-12T13:00:00" -track: "ofbiz" -presenters: "Mike Bates" ---- - -Going back well before the ASF was formed, community has played a vital role in the development of open source software. With the formation of the ASF, we see a great example of open source community being formalized and actively facilitated. Although it can seem counterintuitive, especially to the uninitiated, placing 'Community over code' helps ensure the ongoing relevance of open source as a movement, inviting individuals and organizations alike to invest in 'free' software. The community participation statistics across the ASF are impressive and highly convincing.nThe results of this community participation include personal growth, an ever evolving and strengthening Apache project, as well as a strong foundation for commercial business development.n- Brief historyn- ASF stats overviewn- What 'Community over code' means to various Apache community members At the individual project level, Apache OFBiz has benefitted from hundreds of thousands of volunteer hours (over 75,000 from one application development company alone) over the past 13 years, and relies on numerous other Apache projects. The presenter will share examples of one service provider's approach to making community contributions, including those of individuals as well as company-wide initiatives that have worked well over the years. \ No newline at end of file diff --git a/content/sessions/1066.md b/content/sessions/1066.md deleted file mode 100755 index 3fdf2c66b..000000000 --- a/content/sessions/1066.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Protect your Private Data in your Hadoop Clusters with ORC Column Encryption" -date: "2019-09-11T14:15:00" -track: "bigdata" -presenters: "Owen O'Malley" ---- - -Fine-grained data protection at a column level in data lake environments has become a mandatory requirement to demonstrate compliance with multiple local and international regulations across many industries today. ORC is a self-describing type-aware columnar file format designed for Hadoop workloads that provides optimized streaming reads, but with integrated support for finding required rows quickly. In this talk, we will outline the progress made in Apache community for adding fine-grained column level encryption natively into ORC format that will also provide capabilities to mask or redact data on write while protecting sensitive column metadata such as statistics to avoid information leakage. The column encryption capabilities will be fully compatible with Hadoop Key Management Server (KMS) and use the KMS to manage master keys providing the additional flexibility to use and manage keys per column centrally. An end to end scenario that demonstrates how this capability can be leveraged will be also demonstrated. \ No newline at end of file diff --git a/content/sessions/1067.md b/content/sessions/1067.md deleted file mode 100755 index ba97e4446..000000000 --- a/content/sessions/1067.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Running an Apache Project: 10 Traps and How to Avoid Them" -date: "2019-09-12T16:30:00" -track: "community" -presenters: "Owen O'Malley" ---- - -When you are starting on your open source adventure, there are lot ofnthings to learn that have very little to do with coding and insteadnrelate to interacting with people. Apache is, at its best, a group ofnpeople who are trying to share their experience and teach new projectsnand contributors how to successfully manage open source projects.nHowever, like the blind people each describing a part of an elephant,neach mentor brings their personal experience to the table, and thusncan give good, yet conflicting advice to new projects. However, thatnaggregate advice has helped many projects to become successful. Based on the author's experience, this talk will take you through 10ncommon traps in running Apache projects, why they happen, and how tonavoid or mitigate them. \ No newline at end of file diff --git a/content/sessions/1068.md b/content/sessions/1068.md deleted file mode 100755 index 91402e59a..000000000 --- a/content/sessions/1068.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Using the TLP toolchain as a crystal ball for your cluster" -date: "2019-09-11T11:00:00" -track: "cassandra" -presenters: "Anthony Grasso" ---- - -Cassandra cluster management is hard. Understanding how your Cassandra data model will hold up in production over a period of time can be tricky. If that is not tough enough, understanding how a change to a Cassandra setting will affect your cluster can be be difficult. Knowing how your data model or setting change will perform under a production data load can prevent performance degradation or worse nodes going down. In the last year, The Last Pickle has invested the time to develop the tooling necessary to create test clusters in AWS, as well as a scalable stress tool which can run pre-configured workloads. These tools are designed to take the guess work out of data modelling and configuration changes in Cassandra. In this talk we will introduce our toolchain, look at what each tool does, how they work, and where you would use them. In addition, we will show how the toolchain can be used to quickly test a data model or feature in Cassandra. \ No newline at end of file diff --git a/content/sessions/1069.md b/content/sessions/1069.md deleted file mode 100755 index 00e5312ec..000000000 --- a/content/sessions/1069.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "(Apache) Drill-ing into Collections of HDF5 Files" -date: "2019-09-11T15:15:00" -track: "drill" -presenters: "Hyo-Kyung Joe Lee" ---- - -This talk is about building bridges between two ecosystems, Apache and HDF5.nHDF5 is a widely used storage container for complex data from embedded devicesnto supercomputers, and is developed and maintained as FOSS by The HDF Group.nWhile it is easy to imagine the potential benefits of making HDF5 containersnaccessible from the various Apache frameworks, there are several technicalnchallenges to overcome, and what makes a 'good' integration is by no meansnobvious. It requires the input and collaboration of experts from both sides. The core of this presentation will be an overview of a joint project betwee The HDF Group and Apache Drill contributor Charles Givre. We will show how tonuse Apache Drill to explore collections of HDF5 containers and discuss thenunderlying design decisions and a few technicalities. We will use thisnopportunity to give the Apache community an update on the support forncollections of HDF5 files on HDFS, in cloud storage (such as S3), our Sparkndata source and our new 'HDF5 in the cloud' platform, HDF KITA. The intended audience are 'bridge-builders' and curious members of the Apachencommunity, and anyone interested in 'HDF5 demystified.' \ No newline at end of file diff --git a/content/sessions/1070.md b/content/sessions/1070.md deleted file mode 100755 index 1c8dba428..000000000 --- a/content/sessions/1070.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Consistent Cassandra schema changes in Elassandra" -date: "2019-09-11T17:45:00" -track: "cassandra" -presenters: "Vincent Royer" ---- - -As described in CASSANDRA-10699 (Make schema alterations strongly consistent), concurrent schema changes can still lead to schema disagreement in Cassandra 3.0. In order to properly support Elasticsearch dynamic mapping in Elassandra, we will see how multiple schema changes are validated on a working copy of the Cassandra schema, and applied in an atomic update to all nodes if a light weight transaction succeed, thus avoiding concurrent schema changes issue. I will also explain how we have taken advantage of Cassandra table extensions to store the Elasticsearch mapping directly into the CQL schema with several benefits. \ No newline at end of file diff --git a/content/sessions/1071.md b/content/sessions/1071.md deleted file mode 100755 index 6c8a5e6d9..000000000 --- a/content/sessions/1071.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Diversity, a beginners introduction" -date: "2019-09-10T15:30:00" -track: "community" -presenters: "Nick Burch" ---- - -Diversity and inclusion are becoming hot topics within our open source communities and industry, but for those of us 'playing life on easy', the subject and impact can be harder to 'get'. Like many people who used to play life on easy, I've had some blind spots around diversity and inclusion. I only discovered these through working on projects, communities and support at the ASF, and by moving to a new country! Now it is my mission to help others see the light, learn for themselves, and through that build better open source communities. I'll be sharing what I learnt, giving resources, and specific takeaway actions. \ No newline at end of file diff --git a/content/sessions/1072.md b/content/sessions/1072.md deleted file mode 100755 index 0c2be867c..000000000 --- a/content/sessions/1072.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Evaluating Content/Text Extraction at Scale with Apache Tika" -date: "2019-09-09T17:30:00" -track: "bigdata" -presenters: "Tim Allison" ---- - -Text extraction tools are essential for obtaining the textual content and metadata of computer files for use in a wide variety of applications, including search and natural language processing tools. Techniques and tools for evaluating text extraction tools are largely missing from academia and industry. This talk will focus on recent improvements to Apache Tika's tika-eval module to help integrators evaluate content extraction at scale. The tika-eval module was initially developed for a single batch mode on a single vm. In recent months, however, Apache Tika has refactored this module to allow for easier scaling within Apache Solr and other large scale processing frameworks. This talk will offer an overview of the techniques used to identify potential extraction problems -- including garbled text without ground truth; and the talk will show how the tika-eval module can be used at scale to identify potential problems with content extraction. \ No newline at end of file diff --git a/content/sessions/1073.md b/content/sessions/1073.md deleted file mode 100755 index f5aa3a6c3..000000000 --- a/content/sessions/1073.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Reverse Proxy Magic with Apache httpd 2.4" -date: "2019-09-12T15:30:00" -track: "general" -presenters: "Jim Jagielski" ---- - -One of the best not-so-well-known features of Apache httpd is its incredible capability as a Reverse Proxy server. Whether as a front end to application servers such as Tomcat and PMP-FPM, or as a caching server for static content, or as a robust load-balancer with dynamic configuration, or all, Apache httpd is not only feature rich but easily matches or surpasses the performance of other proxies. \ No newline at end of file diff --git a/content/sessions/1074.md b/content/sessions/1074.md deleted file mode 100755 index aa4324a7d..000000000 --- a/content/sessions/1074.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Cassandra at Instagram 2019" -date: "2019-09-11T14:15:00" -track: "cassandra" -presenters: "Dikang Gu" ---- - -At Instagram, our mission is to capture and share the world's moments. Our app is used by over 1 billion people monthly; this creates a lot of challenging data needs. We use Cassandra behind a lot of different product features. In this presentation, I will talk about how we use Cassandra to serve our critical use cases; the improvements we made in last several years to make sure Cassandra can meet our low latency, high scalability requirements; and future plans. \ No newline at end of file diff --git a/content/sessions/1075.md b/content/sessions/1075.md deleted file mode 100755 index a62803c36..000000000 --- a/content/sessions/1075.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Building and running cloud-native Cassandra" -date: "2019-09-10T18:00:00" -track: "cassandra" -presenters: "Vinay Chella, Joey Lynch" ---- - -This session will re-evaluate Cassandra’s relationship with runtime and build systems, pointing out ways that the existing systems fall down, and identifying avenues for improvement. Over the past few years, a number of platforms have emerged for running user code. Container runtimes like Docker, container orchestrators such as Kubernetes, and metrics collections agents like Prometheus and Spectator have all gained popularity and mind-share. Cassandra functionality such as metrics, bootstrapping, and monitoring integrates with the newer paradigms, but in an ad-hoc and improvised fashion. By taking a purposeful approach to integrating with these new methods of deployment, the Cassandra community can more fully benefit from their advertised strengths. The Cassandra build system based on Ant+Ivy dates to the early 2000’s, and reflects legacy complexity that could be avoided with modern build systems. Cassandra’s system package builds are not much better and often fail to integrate with industry standards such as systemd. Iterating on the existing systems is difficult, but this technical debt slows innovation in our build systems. In this talk, we propose solutions to make building, deploying and monitoring Cassandra easy and low overhead, while taking advantage of cloud advancements wherever possible. \ No newline at end of file diff --git a/content/sessions/1076.md b/content/sessions/1076.md deleted file mode 100755 index df9fa8ae0..000000000 --- a/content/sessions/1076.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache Cassandra community health" -date: "2019-09-10T11:15:00" -track: "cassandra" -presenters: "Ben Bromhead, Scott Andreas" ---- - -This talk will discuss the current health of the Apache Cassandra community looking at everything from community adoption, user adoption, health of contributors/committers, project velocity and a number of other measures of community health. In this talk Ben will also discuss broader market changes and challenges that impact the Apache Cassandra community drawn from our experience at Instaclustr, discussions with industry analysts and public data sources. This talk will be an update on the state of Apache Cassandra community presentation from NGCC 2017. \ No newline at end of file diff --git a/content/sessions/1077.md b/content/sessions/1077.md deleted file mode 100755 index 7d2e79d21..000000000 --- a/content/sessions/1077.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Self-Service IoT Data Analytics with StreamPipes" -date: "2019-09-10T15:30:00" -track: "internetofthings" -presenters: "Dominik Riemer" ---- - -In many application domains such as manufacturing, the integration and continuous processing of real-time sensor data from the Internet of Things (IoT) is crucial to continuously monitor and detect upcoming situations. While continuous processing of events in scalable architectures is already well supported by the existing Big Data tool landscape, building such applications requires technical effort and programming skills, which is often not present in manufacturing companies. To solve this problem, we have developed StreamPipes (https://www.streampipes.org), which aims at simplifying IoT data analytics for domain experts. Originating from a research project, StreamPipes has been publicly released under the Apache License in early 2018. At its core, StreamPipes provides non-technical users with an easy and intuitive way to create stream processing pipelines through a graphical editor. Pipelines consist of reusable pipeline elements, each of them implemented as a microservice using a wrapper for a specific Big Data technology (e.g., Apache Flink, Kafka Streams or lightweight runtimes that can run directly at the edge). We motivate our talk by showing examples we gathered from a number of industry projects during the past years in Industrial IoT domains, present technical details of StreamPipes and show how the tool eases the accessibility of Big Data tools for non-technical users. \ No newline at end of file diff --git a/content/sessions/1078.md b/content/sessions/1078.md deleted file mode 100755 index c4df8c998..000000000 --- a/content/sessions/1078.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Hello World! Introducing Apache IoTDB - a Database for the Internet of Things" -date: "2019-09-10T17:00:00" -track: "internetofthings" -presenters: "Xiangdong Huang, Julian Feinauer" ---- - -Time series data management system is becoming increasingly important with the rise of the Internet of Things, especially in the industrial area. Apache IoTDB (incubating), one of the youngest projects in the Apache Incubator, is born for that!nApache IoTDB is an integrated data management engine designed for time series data, based on a file format which is able to store time series data very efficiently. It provides users specific services for data collection, storage and analysis. Due to its lightweight structure, high performance and usable features together with its good integration with Apache Hadoop and Apache Spark, it can meet the requirements of massive dataset storage, high-speed data ingestion and complex data analysis for industrial IoT applications.nAs the project is very new to the ASF, the talk has four objectives. First, discuss the time series data management requirement in the IoT area; Second, introduce Apache IoTDB to the world; Third, to give an overview about the architecture of the project and the technologies used inside and Fourth, to give an idea about how IoTDB integrates with other Apache Projects and where it fits in the (IoT) data processing ecosystem. Some practical use cases are also introduced. \ No newline at end of file diff --git a/content/sessions/1079.md b/content/sessions/1079.md deleted file mode 100755 index 580f5376f..000000000 --- a/content/sessions/1079.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Episode IV - A New Hope: Industry 4.0 done our way" -date: "2019-09-09T11:30:00" -track: "internetofthings" -presenters: "Christofer Dutz" ---- - -It is a period of technological change. Rebel Industry 4.0 open-source software solutions, striking from hidden bases, have won their first victories against the evil proprietary Industrial Empire. During the battle, Rebel spies managed to implement industrial protocols to the Empire’s ultimate weapons, the industrial Programmable Logic Controllers, almost unprotected pieces of hardware with enough power to destroy an entire planet. Pursued by the Empire’s sinister agents, a representative of the project races to Las Vegas aboard his intercontinental sub-orbital shuttle, custodian of his plans to revolutionize the way we are building Industry 4.0 applications, that can free all companies and restore freedom to the industry … Learn how the Apache PLC4X project has evolved and what we have planned for the future and especially how it can currently help close the gap between industrial hardware and modern software solutions to build the next generation of industrial IoT solutions, based entirely on open-source software. \ No newline at end of file diff --git a/content/sessions/1080.md b/content/sessions/1080.md deleted file mode 100755 index f3d6c8def..000000000 --- a/content/sessions/1080.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "SPARQL at scale with Apache Rya" -date: "2019-09-11T14:15:00" -track: "graphprocessing" -presenters: "Adina Crainiceanu" ---- - -The Resource Description Framework (RDF) is a standard model for storing graph data. While the standard was initially created for storing meta-data about the World Wide Web, its flexible format made it a popular choice for storing many different types of information. With the explosive increase in the size of available data, scalable solutions are needed to efficiently store and query very large RDF graphs within big data architectures. Apache Rya (incubating) is a scalable database management system designed for storing and searching very large RDF data. Rya is built on top of Apache Accumulo and also supports a MongoDB back end. nIn this talk, we introduce storage methods, primary and secondary indexing schemes, statistics based query optimization, as well as query evaluation techniques that allow Rya to scale to billions of triples across multiple nodes, while providing fast and easy access to the data through conventional query mechanisms such as SPARQL. \ No newline at end of file diff --git a/content/sessions/1081.md b/content/sessions/1081.md deleted file mode 100755 index 6f2a45871..000000000 --- a/content/sessions/1081.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Standardizing Industrial IoT with Apache PLC4X, Apache Daffodil, and DFDL" -date: "2019-09-10T18:00:00" -track: "internetofthings" -presenters: "Steve Lawrence" ---- - -Industrial IoT is made up of many different programmable logic controllers (PLCs) that communicate with often incompatible and proprietary data formats. Apache PLC4X standardizes access to these PLCs by creating a set of libraries across a variety of languages to read and write PLC data with a unified API. Unfortunately, developing and maintaining support for these different data formats, especially across multiple languages, can be burdensome. In this talk we discuss how we can describe PLC data using the open standard Data Format Description Language (DFDL), how Apache Daffodil uses these descriptions to transform data to a standard and easily consumable representation, and an approach to extend PLC4X to use Daffodil and DFDL descriptions to reduce the effort needed to support a variety of PLCs and languages. \ No newline at end of file diff --git a/content/sessions/1082.md b/content/sessions/1082.md deleted file mode 100755 index 32d0538a5..000000000 --- a/content/sessions/1082.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Everything about Apache Marvin Platform" -date: "2019-09-12T15:30:00" -track: "machinelearning" -presenters: "Daniel Takabayashi" ---- - -Apache Marvin is an open source project that focuses on empowering data science teams to deliver industrial-grade applications supported by a high-scale, low-latency, language agnostic, and standardized architecture platform created to simplify the process of exploration and modeling in Machine Learning projects. Marvin aims at abstracting the complexities in the creation process of scalable, highly available, interoperable and maintainable predictive software. In this presentation, the speaker is going to talk about the architecture behind the platform and show the cooler features of Apache Marvin. \ No newline at end of file diff --git a/content/sessions/1083.md b/content/sessions/1083.md deleted file mode 100755 index 5ac7cf8c2..000000000 --- a/content/sessions/1083.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Tomcat from a cluster to a cloud." -date: "2019-09-10T18:00:00" -track: "tomcat" -presenters: "Jean-Frederic Clere" ---- - -Using Tomcat in a cluster and in a cloud. nWe start by looking how to configure tomcat to get a cluster andnthen explore the problems and solutions to have distributed applications running in a cloud. Most cloud providers now have a Kubernetes API.nWe will look to what we have to add to Tomcat to have a decent cloud supportnfor monitoring, tracing and operating on the cloud.nWe will show how to use all the pieces.nA demo of a cluster will be prepared and run during the presentation andnthe corresponding application will be moved to a Kubernetes cloud. \ No newline at end of file diff --git a/content/sessions/1084.md b/content/sessions/1084.md deleted file mode 100755 index dd126fa88..000000000 --- a/content/sessions/1084.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "How to Grow Communities, Not Just Code" -date: "2019-09-10T18:00:00" -track: "community" -presenters: "Ruth Suehle" ---- - -Good code isn't enough for a successful open source project. First of all, only you know how to use what you've made. Maybe it's time for a little UI and UX help? At the very least some documentation! Next, how is anyone else going to find what you've created? And that's only the beginning. Ruth Suehle, manager of Red Hat's Open Source and Standards community leadership team, will take you through examples of the best and the worst, from projects large and small, to help you see what you need beyond your code to build a successful open source project and community. \ No newline at end of file diff --git a/content/sessions/1085.md b/content/sessions/1085.md deleted file mode 100755 index 991cf3361..000000000 --- a/content/sessions/1085.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "IoT Applications and Patterns using Apache Spark & Apache Bahir" -date: "2019-09-10T14:30:00" -track: "internetofthings" -presenters: "Luciano Resende, Leucir Marin" ---- - -The Internet of Things (IoT) is all about connected devices that produce and exchange data, and building applications that produce insights from these high volumes of data are very challenging and require an understanding of multiple protocols, platforms, and other components. On this session, we will start by providing a quick introduction to IoT, some of the common analytic patterns used on IoT, and also touch on the MQTT protocol and how it is used by IoT solutions some of the quality of services tradeoffs to be considered when building an IoT application. We will also discuss some of the Apache Spark platform components, the ones utilized by IoT applications to process devices streaming data. We will also talk about Apache Bahir and some of its IoT connectors available for the Apache Spark platform. We will also go over the details on how to build, test and deploy an IoT application for Apache Spark using the MQTT data source for the new Apache Spark Structure Streaming functionality. \ No newline at end of file diff --git a/content/sessions/1086.md b/content/sessions/1086.md deleted file mode 100755 index 42ec6233b..000000000 --- a/content/sessions/1086.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Adding AI to customer segmentation using Apache Unomi and Apache PredictionIO" -date: "2019-09-12T15:30:00" -track: "machinelearning" -presenters: "Serge Huber" ---- - -Apache Unomi is a Customer Data Platform that is easy to extend as it uses the Apache Karaf runtime. In this session, you will discover how to integrate Unomi with Apache PredictionIO to be able to leverage artificial intelligence algorithms to recommend segments or products to users as they interact with websites. Of course, as all of this will be bleeding-edge stuff, expect plenty of last-minute drama, half-finished code and hopefully some fun demos. \ No newline at end of file diff --git a/content/sessions/1087.md b/content/sessions/1087.md deleted file mode 100755 index a6788c669..000000000 --- a/content/sessions/1087.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Proxies, TLS, and Injecting Security Into Cloud Communication" -date: "2019-09-10T15:30:00" -track: "contentdelivery" -presenters: "Susan Hinrichs" ---- - -With the expansion of moving legacy applications into public clouds, there have been an increased need for features in proxies to inject security into legacy application deployments. In this talk, we present extensions made to the Apache Traffic Server proxy over the past two years to enable the secure deployment of legacy applications in a wider variety of network configurations. Specifically, we discuss how Traffic Sever can be used for Layer 4 Routing. In this case, the router does not terminate the TLS connection, but rather use the Server Name Indicator (SNI) to determine were the TLS connection should actually be terminated. Traffic Server configuration additions were also made to allow for the finer-grained deployment of mutual TLS authentication both between the user agent and Traffic Server and between Traffic Server and the origin. A new TLS Bridge plugin allows for the injection of a proxy-to-proxy TLS tunnels to support older applications securely communicating between organizations (e.g. between public cloud and on premise cloud). While we have specifically been working with Traffic Server, our development and production deployment experiences for moving applications into new cross-cloud configurations apply to the problem space in general. \ No newline at end of file diff --git a/content/sessions/1088.md b/content/sessions/1088.md deleted file mode 100755 index aef4e2f46..000000000 --- a/content/sessions/1088.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Robust Elastic Scaling of Deep Learning on Kubernates" -date: "2019-09-12T16:30:00" -track: "machinelearning" -presenters: "Junfeng Liu" ---- - -With the increasing popularity of deep learning workloads in academia, government and industry, several organizations have developed cloud-based deep learning platforms whose goals are to enable organizations utilize expensive resources effectively, and to share said resources among multiple teams in a fair and effective manner. We present the elastic scaling in the context of such large scale deep learning platforms. The traditional deep learning workload is the parallel job where can not fully utilize the cluster gpu resources. We begin by analyzing deep learning training workloads, and how they differ from conventional workloads as far as elastic scaling is concerned. The scheduler driven scaling to optimize resource utilization, support priorities, maintenance, at the same time promising the accuracy and user can define the strategy for scale out. The presentation will describe the design and implementation of a DL workload aware scheduling supporting flexible scheduling-driven scaling on the top of kubernetes cluster. Compare with kub-batch, kub-flow, the new scheduler offer much better job through put. \ No newline at end of file diff --git a/content/sessions/1089.md b/content/sessions/1089.md deleted file mode 100755 index 3bd73b67a..000000000 --- a/content/sessions/1089.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache Toree: A Jupyter Kernel for Scala / Apache Spark" -date: "2019-09-12T10:00:00" -track: "machinelearning" -presenters: "Luciano Resende" ---- - -Many data scientists are already making heavy usage of the Jupyter ecosystem for analyzing data using interactive notebooks. Apache Toree (incubating) is a Jupyter kernel designed that enables data scientists and data engineers to easily connect and leverage Apache Spark and its powerful APIs from a standard Jupyter notebook to execute their analytics workloads. In this talk, we will go over what's new with the most recent Apache Toree release. We will cover available magics and visualizations extensions that can be integrated with Toree to enable better data exploration and data visualizations. We will also describe some high-level design of Toree and how users can extend the functionality of Apache Toree powerful plugin system. And all of these with multiple live demos that demonstrate how Toree can help with your analytics workloads in an Apache Spark environment. \ No newline at end of file diff --git a/content/sessions/1090.md b/content/sessions/1090.md deleted file mode 100755 index 09cc2c823..000000000 --- a/content/sessions/1090.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Behind the Scenes of the Apache Software Foundation" -date: "2019-09-10T14:30:00" -track: "community" -presenters: "Lars Eilebrecht" ---- - -This presentation will give you everything you always wanted to know about the Apache Software Foundation (ASF), but were afraid to ask. It will show you that there is more than just the Apache web server, and provide you with information on how the ASF works and its history. The difference between membership and committership, who decides what, how elections take place, the technical infrastructure, project management committees, and the philosophy behind the incubator. Come and see behind the scenes of the Apache Software Foundation and let's talk about 'The Apache Way' and 'Community over Code' which have been the cornerstones of the ASF for 20 years. \ No newline at end of file diff --git a/content/sessions/1091.md b/content/sessions/1091.md deleted file mode 100755 index f0c607ebf..000000000 --- a/content/sessions/1091.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "A Beginner's Guide to HTTPS and TLS" -date: "2019-09-12T16:30:00" -track: "general" -presenters: "Lars Eilebrecht" ---- - -This talk will introduce you to the fundamentals of securing communications with HTTPS. We will start by explaining the basics of X.509 server and client certificates, certification authorities, and using the OpenSSL toolkit. The TLS/SSL protocol will be introduced and how it is used together with HTTP in order to provide for data encryption, integrity, and authentication. We will talk through some important configuration details, standard use cases, common pitfalls, known SSL vulnerabilities, and issues when using HTTPS. The Apache HTTP Server will be used to provide specific examples, but the general information of the talk will be applicable to most server software that supports HTTPS. \ No newline at end of file diff --git a/content/sessions/1092.md b/content/sessions/1092.md deleted file mode 100755 index 07b60a3eb..000000000 --- a/content/sessions/1092.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "How China Search Giant Baidu adopt apache way inside its company" -date: "2019-09-12T10:00:00" -track: "community" -presenters: "Tanzhongyi" ---- - - Baidu is one of the biggest Internet company, it was founded 20 years ago with 10,000 more engineers now.nit began to embraces open source recently years, including adopting InnerSource and contributing to open source community. InnerSource is the use of apache ways inside the company. n As the leader of this program, I will talk about how this happened, and what's the challenges we faced and overcome. we need to set policies, define processes and enable tools , but the most important part is to cultivate open source culture, educate engineers what is apache way and how to cooperate inside company just like in open source community. And now, many projects inside Baidu have been adopting InnerSource, and some projects have became mature and open sourced to external world, even contributed to open source foundation.nWith my affords, Baidu has contributed three projects into apache software foundation as incubator projects,nthere are echarts, Doris, brpc. But open source is a long journey, we need to insist and keep going. \ No newline at end of file diff --git a/content/sessions/1093.md b/content/sessions/1093.md deleted file mode 100755 index 0998bdc96..000000000 --- a/content/sessions/1093.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Cassandra: Partition Management" -date: "2019-09-10T15:30:00" -track: "cassandra" -presenters: "Matija Gobec, Asias He" ---- - -(Cassandra NGCC) Apache Cassandra on-disk storage model is based on immutable SSTables that are, depending on the compaction strategy, optimized for some general use cases. The main issue with optimally reading on-disk data is that a single partition data most of the time needs to be fetched from multiple partitions. None of the existing compaction strategies are optimized to provide a single file read guarantees on cold data. Partition based compaction strategy addresses that issue with effectively compacting all data that belongs to a single partition into the same SSTable. In this presentation, we will talk about proposed implementation details for such compaction strategy and a couple of solutions addressing the issues like write amplification and higher disk io requirements compared to default size tiered compaction. \ No newline at end of file diff --git a/content/sessions/1094.md b/content/sessions/1094.md deleted file mode 100755 index 6f1d004ed..000000000 --- a/content/sessions/1094.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Using Apache MXNet to Train and Deploy your Deep Learning model" -date: "2019-09-12T13:00:00" -track: "machinelearning" -presenters: "Qing Lan" ---- - -AI is evolving rapidly, and much of the recent advancement is driven by Deep Learning, a machine learning technique inspired by the inner-working of the human brain. In this session, we will discuss what deep learning is, and the new capabilities it enables. We will dive into a few computer vision and natural language models that are demonstrating super-human performance, and to integrate these models into your existing Java system leveraging Apache MXNet - an open source deep learning framework – and it's Java API. In our session, we will show a brief introduction to DL, specifically about applications using MXNet Gluon to do the training and Java to do inference with different DL models in Production systems. We would also demostrate how to run and deploy the DL models with MXNet Java API. By the end of the session, audience will learn how to leverage deep learning tasks using Apache MXNet, the various gotchas involved, and where/how to learn more. \ No newline at end of file diff --git a/content/sessions/1095.md b/content/sessions/1095.md deleted file mode 100755 index b65d9266d..000000000 --- a/content/sessions/1095.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Cassandra 4.0: Our Most Stable Major Release" -date: "2019-09-12T16:30:00" -track: "cassandra" -presenters: "Jordan West" ---- - -In mid-2018 the Cassandra community committed to making Cassandra 4.0 the most stable major release of Cassandra in the project’s history. In September, the community shifted focus from feature work and development to ensuring the quality of the release. To this end, we have adopted several new approaches to testing and validation including the replaying of production traffic, code audits, and property-based testing. This talk will explore the methodologies we’ve adopted and the results of their application as well as costs of this level of commitment to testing and its benefits. \ No newline at end of file diff --git a/content/sessions/1096.md b/content/sessions/1096.md deleted file mode 100755 index 0bda04255..000000000 --- a/content/sessions/1096.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "How to Slide Your Release Pass the Incubator" -date: "2019-09-12T09:00:00" -track: "community" -presenters: "Justin Mclean" ---- - -All podling releases need to be voted on by the incubator PMC before being released to the world. I'll go through what the incubator PMC looks for in every release and what you can do to make it pass that IPMC vote and get your project one step closer to graduation. More importantly I'll cover where you can get help if you need it. In this talk, I'll describe current incubator and ASF policy, recent changes that you may not be aware of, and go into detail the legal requirements of common open source licenses and the best way to assemble your NOTICE and LICENSE files. Where possible I describe the reasons behind why things are done a certain which may not always be obvious from our documentation. I'll show how I review a release and the simple tools I use. I'll go through a worked example or two, including a fictional project called Apache Wombat, and cover common mistakes I've seen in releases. \ No newline at end of file diff --git a/content/sessions/1097.md b/content/sessions/1097.md deleted file mode 100755 index 085ddd66a..000000000 --- a/content/sessions/1097.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Openstack from a CloudStack perspective" -date: "2019-09-09T11:30:00" -track: "cloudstack" -presenters: "Rafael Weingärtner" ---- - -An experienced CloudStack developer will show his journey to deploy an OpenStack cloud. We will compare the pros and cons of CloudStack and OpenStack from the following perspectives:n(i) Deployment: cost, time and risks of deploying a CloudStack vs OpenStack infrastructure;n(ii) Maintenance: the costs and time required on maintaining a Cloud orchestrated with CloudStack and another with OpenStack; and,n(ii) Development: how development on both orchestrators is done? What looks like the codebase of each project and which technologies are involved? What kind of professional/development team is required? \ No newline at end of file diff --git a/content/sessions/1098.md b/content/sessions/1098.md deleted file mode 100755 index a2dfd5321..000000000 --- a/content/sessions/1098.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Troubleshooting the Virtual Router - Run and Get Diagnostics" -date: "2019-09-09T14:00:00" -track: "cloudstack" -presenters: "Boris Stoyanov - aka Bobby" ---- - -Demonstration of CloudStack’s latest features for troubleshooting the Virtual Router connectivity and configuration, called Run and Get Diagnostics. Run Diagnostics gives the admin ability to execute diagnostics commands native the VR OS directly from the CloudStack UI. Admin is able to determine connectivity capabilities of the VR without logging into the SystemVM at all. Get diagnostics feature allow the admin to gather information about the SystemVM as fast as clicking on a button in the ‘QuickView’ of the VM. Admin can also execute custom scripts on each SystemVM type from the CloudStack UI. This session will begin with complete presentation of the features followed by live demo and Q&A. \ No newline at end of file diff --git a/content/sessions/1099.md b/content/sessions/1099.md deleted file mode 100755 index 4b4c4affc..000000000 --- a/content/sessions/1099.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache Kibble Crash Course" -date: "2019-09-11T16:45:00" -track: "community" -presenters: "Daniel Gruno" ---- - -This presentation aims at showcasing the Apache Kibble project, getting new users set up with their own Kibble instance and getting started with project and community metrics. During this talk, the audience will learn what Apache Kibble is, what it does, HOW it does it, and how you can easily get set up in a few minutes and start reviewing metrics from your own project. We will also show a demo instance with a select handful of Apache projects, and dive into both single community metrics as well as comparative measuring across foundations. This presentation is aimed at both audiences wishing to start using community metrics and KPIs on their own hardware, as well as people just along for a quick demo of the project. The presentation will end with a discussion about KPIs and areas where these metrics can improve. \ No newline at end of file diff --git a/content/sessions/1100.md b/content/sessions/1100.md deleted file mode 100755 index da6516bf3..000000000 --- a/content/sessions/1100.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "State of Storage in Cloudstack" -date: "2019-09-09T15:00:00" -track: "cloudstack" -presenters: "Sid Kattoju" ---- - -With the increased focus and legislation around data sovereignty requirements, storage solutions play an increasingly important role in cloud infrastructures. We will present an overview of the evolution of storage orchestration in Apache CloudStack, with a focus on managed storage. It will also shed light on some upcoming features and the future direction. \ No newline at end of file diff --git a/content/sessions/1101.md b/content/sessions/1101.md deleted file mode 100755 index 0d78d651d..000000000 --- a/content/sessions/1101.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Running C* on Kubernetes with CassKop, the C* operator for K8S developed by Orange" -date: "2019-09-12T10:00:00" -track: "cassandra" -presenters: "Jean Armel Luce, Sebastien Allamand" ---- - -This presentation will describe the initial experience building and using CassKop, an operator developed for running Cassandra on top of Kubernetes. CassKop works as a usual K8S controller (reconcile the real state with a desired state) and automates the Cassandra operations through JMX. All the operations are launched by calling standard K8S APIs (kubectl apply …) or by using a K8S plugin (kubectl casskop …). CassKop is developed in GO, based on CoreOS operator-sdk framework.nAmong the main features :n- deploying a rack aware cluster (or AZ aware cluster)n- scaling up & down (including cleanups)n- setting and modifying configuration parameters (C* and JVM parameters)n- adding / removing a datacenter in Cassandran- rebuilding nodesn- removing node or replacing node (in case of hardware failure)n- upgrading C* or Java versions (including upgradesstables)n- monitoring (using Prometheus/Grafana)n- ... By using local and persistent volumes, it is possible to handle failures or stop/start nodes for maintenance operations with no transfer of data between nodes. Moreover, cassandra-reaper is deployed in K8S and used for scheduling repair sessions (Thanks to Spotify and TheLastPickle teams)nThe Cassandra exporter for Prometheus and backup/restore developed by Instaclustr are also used (Thanks to Instaclustr team) During this session, we will delve into the architecture and implementation of this operator, and share our learnings. \ No newline at end of file diff --git a/content/sessions/1102.md b/content/sessions/1102.md deleted file mode 100755 index a08c17024..000000000 --- a/content/sessions/1102.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Modernizing the Nordstrom Supply Chain: Efficient Order Fulfillment using JanusGraph, Cassandra and Spark" -date: "2019-09-11T12:00:00" -track: "graphprocessing" -presenters: "Jeff Callahan" ---- - -DESCRIPTION: Today's Nordstrom customers expect order fulfillment options shipping to the most convenient locations in the timeframes that best fit their schedules. This session offers an inside look at how Nordstrom's Supply Chain Technology team combined an extensive network of brick and mortar facilities with a custom software stack built atop JanusGraph to present fulfillment options tailored to each individual order. ABSTRACT: Over the last decade, the volume of online retail sales as a proportion of overall retail sales has nearly tripled, and there are no signs of this trend slowing down. In a $500 billion online retail market driven by ever more savvy online shoppers, expectations for order fulfillment have grown more demanding. Consumers expect a choice of fulfillment options that offer convenience, flexibility and value. Nordstrom's Supply Chain Technology team is charged with meeting those customer expectations through efficient distribution of millions of unique items from a cross-country network of brick and mortar facilities. For the engineers on our team, it made perfect sense to use JanusGraph to model a dynamic inventory picture managed by a diverse set of geographically distributed facilities. Despite our enthusiasm, not everyone was eager to adopt an unfamiliar open source graph system. In this presentation, Jeff Callahan will discuss why his team at Nordstrom embraced JanusGraph as the backbone of a new data pipeline built to meet the evolving expectations of customers today (and tomorrow, too). After taking a detailed look at the system architecture, we'll also discuss some of the internal organizational challenges we faced in adopting JanusGraph, and how we managed to navigate those challenges to ultimately deliver an even stronger solution. NOTE: While this presentation focuses on JanusGraph, there is considerable overlap with Big Data, Machine Learning and Cassandra \ No newline at end of file diff --git a/content/sessions/1103.md b/content/sessions/1103.md deleted file mode 100755 index cc8fc47ec..000000000 --- a/content/sessions/1103.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Conquering Networked Applications using Ballerina Programming Language" -date: "2019-09-12T13:00:00" -track: "integration" -presenters: "Anjana Fernando" ---- - -Ballerina is the next generation programming language, which redefined what means to be 'general purpose'. Historically, programming languages concentrated on single machine execution, in a controlled environment, and also for good reason, because any other external interactions were out of the scope of a programming language, at least in the bygone days that is. But now, the communication network is something that is always there, and software often doesn't work alone. But rather, they work more and more by communicating with each other to do something meaningful. So hiding this network from our code is no longer an option, and we try to use various frameworks and libraries get the functionality we need. Ballerina is a general purpose programming language, which has built its core concepts and functionalities to support creation of networked application. These features include the built-in support for services/resources, transactions, and resilient communication support, and also combines with a type system which further enhances these operations. In this session, we will go through Ballerina, in understanding the motivation behind this language, the features it introduces, and why it is a critical and a timely addition to the tools and technologies we would need now. \ No newline at end of file diff --git a/content/sessions/1104.md b/content/sessions/1104.md deleted file mode 100755 index 7fc56a34a..000000000 --- a/content/sessions/1104.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Building S3 over Ozone : Making a Cloud Native File System" -date: "2019-09-11T15:15:00" -track: "bigdata" -presenters: "Bharat Viswanadham, Anu Engineer" ---- - -The AWS S3 protocol is the defacto interface for modern object stores. Ozone supports S3 protocol as the first-class notion in Ozone. For all practical purposes, a user of S3 can start using Ozone without any change to code or tools. S3 protocol support offered by Ozone is strongly consistent, so users don’t need to run sidekick tools like S3Guard when running big data applications like Apache Spark, Apache YARN or Apache Hive. This talk is a deep dive into how Ozone supports S3, and how easy it is to use the cloud tools against Ozone. We will demo S3 over Ozone with different off-the-shelf tools like Goofys, Kubernetes CSI, AWS CLI and also demonstrate how to use applications like Apache Spark, Apache Hive and YARN. We will also discuss some of the challenges faced during building S3 compatible rest server. For example, how we did the mapping of S3 semantics to ozone semantics, how we solved the S3 security issues and mapped them to ozone world. With S3 support Ozone has become a generic object store: it can be used with thousands of s3 compatible tools while it will also provide Hadoop compatible file system access. \ No newline at end of file diff --git a/content/sessions/1105.md b/content/sessions/1105.md deleted file mode 100755 index c2000c339..000000000 --- a/content/sessions/1105.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache Mynewt - ASF's Embedded OS Project" -date: "2019-09-09T10:00:00" -track: "internetofthings" -presenters: "Justin Mclean" ---- - -Apache Mynewt is an operating system for low powered embedded systems based on 32-bit microcontrollers. At the core, it is a small pre-emptive RTOS and an extensive set of modules for connectivity, file systems, power and performance management options and lots more. It is a highly flexible, responsive, general purpose OS for constrained devices, think of it as embedded Linux for devices that can’t run Linux. Features include a secure bootloader, image management, hardware abstraction, instrumentation for stats and logs. This makes it easy to go from prototype to production when using Apache Mynewt without having to reinvent the wheel. \ No newline at end of file diff --git a/content/sessions/1106.md b/content/sessions/1106.md deleted file mode 100755 index 1ceb4e86c..000000000 --- a/content/sessions/1106.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache NetBeans clever APIs for cross platform development" -date: "2019-09-12T10:00:00" -track: "general" -presenters: "Anton Epple" ---- - -Hybrid applications can be a great solution for cross platform development, but JavaScript is not everyones cup of coffee. With a small set of clever APIs and SPIs, NetBeans has created a lean cross platform framework, that allows you to code in Java, and run the application on any mobile or desktop platform, as hybrid, or fully native application. In this sessions I'd like to unveil this hidden gem (improperly named 'HTML/Java'), show some real life applications and discuss how this improves client side software development dramatically. \ No newline at end of file diff --git a/content/sessions/1107.md b/content/sessions/1107.md deleted file mode 100755 index c24a50dc8..000000000 --- a/content/sessions/1107.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Processing IoT Data from End to End with MQTT and Apache Kafka" -date: "2019-09-10T12:15:00" -track: "internetofthings" -presenters: "Kai Waehner" ---- - -This session discusses end-to-end use cases such as connected cars, smart home or healthcare sensors where you integrate IoT devices with enterprise IT using open source technologies and standards. MQTT is a lightweight messaging protocol for Internet of Things (IoT). However, MQTT is not built for high scalability, longer storage or easy integration to legacy systems. Apache Kafka is a highly scalable distributed streaming platform, which ingests, stores, processes and forwards high volumes of data from thousands of IoT devices. The session discusses how to leverage the Apache Kafka open source ecosystem as streaming platform to process IoT data. See a live demo of how MQTT brokers like Mosquitto integrate with Kafka, and how you can even integrate MQTT clients to Kafka without MQTT Broker. Learn how to analyze the IoT data either natively on Kafka with Kafka Streams / KSQL or on an external big data cluster like Spark, Flink or Elastic leveraging Kafka Connect. \ No newline at end of file diff --git a/content/sessions/1108.md b/content/sessions/1108.md deleted file mode 100755 index a598100b7..000000000 --- a/content/sessions/1108.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Machine Learning applications for CloudStack" -date: "2019-09-10T15:30:00" -track: "cloudstack" -presenters: "Anurag Awasthi" ---- - -While Machine learning and data mining has had profound impact on how we model applications and use data for better product consumption, there is scope for extending prediction algorithms to lower levels as well. Some useful applications of machine learning in ACS could be exploring better resource allocation that is aware of usage statistics, predicting faults, load balancing, etc. In this talk we will * take a broad overview of what Machine Learning/Data mining is and how it is being used in today's tech ecosystemn* explore ways in which we can make ACS more efficientn* discuss some recent advancements in how ML can benefit datacenters from research community \ No newline at end of file diff --git a/content/sessions/1109.md b/content/sessions/1109.md deleted file mode 100755 index 6dfce4fb8..000000000 --- a/content/sessions/1109.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "NoVNC console implementation in CloudStack" -date: "2019-09-09T11:30:00" -track: "cloudstack" -presenters: "Wei Zhou" ---- - -NoVNC is an open-source browser based VNC client, implemented using HTML5 technologies (Web Sockets, Canvas) with encryption (wss://) support. Proposed NoVNC implementation provides better performance and is more responsive when compared to the current ajax-based console in Apache CloudStack. Our implementation is based on NoVNC and websockify. Websockify is a websocket proxy converting WebSocket request from browser to TCP request and then forward it to VNC server. We have already implemented a NoVNC console for CloudStack 4.13. This new console is smooth and very stable. It can be accessed by multiple sessions at the same time and easily integrated into other platforms. During this talk we will demonstrate how NoVNC console is implemented in CloudStack and how VNC password authentication is implemented in NoVNC and websockify. \ No newline at end of file diff --git a/content/sessions/1110.md b/content/sessions/1110.md deleted file mode 100755 index d6500512b..000000000 --- a/content/sessions/1110.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Mission Critical Cassandra in Financial Services, The When/What/How Experiences for ALL!" -date: "2019-09-11T16:45:00" -track: "cassandra" -presenters: "Jonathan Koppenhofer" ---- - -Apache Cassandra is a database product that is most notably known for being highly scalable and fault tolerant. Since these characteristics are something that is desirable in almost every use case, it attracts attention from developers up through executives when faced with decisions on how to architect mission critical systems. If these traits are so desirable, and Apache Cassandra makes these capabilities available to end users, then shouldn’t we use Cassandra for all our data storage needs? What is the catch? When considering Apache Cassandra for one’s database needs, it is critical for users to understand when it is appropriate for use, and what are the trade-off’s one needs to consider. A person must consider the priorities of their non-functional requirements, the required investment in knowledge acquisition, the increased complexity of distributed systems, and the operational needs of maintaining such a database system among other concerns. If used incorrectly, the results could be disastrous, but when used appropriately Apache Cassandra can produce the expected rewarding results. In this presentation, [redacted name] will share his experiences operating one of the largest Apache Cassandra deployments in the financial services industry. As such, specific topics will be detailed that are relevant to both the largest and smallest of use cases.n* When is it appropriate to use Cassandra? What has been successful/unsuccessful?n* What have been the largest barriers to adoption?n* How to operate at any scalen* Risk and Control considerationsn* Knowledge acquisition and proliferation. Please join us for this candid discussion in hopes of easing the journey to successful adoption of Cassandra. \ No newline at end of file diff --git a/content/sessions/1111.md b/content/sessions/1111.md deleted file mode 100755 index 9cabdce94..000000000 --- a/content/sessions/1111.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Configuring Apache Camel for the Cloud" -date: "2019-09-12T15:30:00" -track: "integration" -presenters: "Bob Paulin" ---- - -Cloud … check. Container … check. Orchestration … check. Now that you’ve got the basics it’s time to starting thinking about how all occupants of your new cloud infrastructure are going to communicate. And no mater what cloud, container or orchestration tool you choose, Apache Camel has what you need to get your system configured and connected. This talk will cover Camel components for cloud friendly configuration, communication, packaging, and deployment. The approaches presented will take a “use what is already there” philosophy. So whatever you’re working with from Do It Yourself Virtual Machines with containers like Apache Karaf or a fully managed Kubernetes clusters with fat jars, integrating vertically or horizontally, Apache Camel has the pieces to make it all work. This talk will cover how to do it as as well as advantages and drawbacks of each approach. \ No newline at end of file diff --git a/content/sessions/1112.md b/content/sessions/1112.md deleted file mode 100755 index 6a85216b0..000000000 --- a/content/sessions/1112.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Welcome to Cloudstack Collaboration Conference" -date: "2019-09-09T09:00:00" -track: "cloudstack" -presenters: "Paul Angus" ---- - -The CloudStack Project VP welcomes delegates to the CloudStack Collaboration Conference. And gives opening remarks about the CloudStack community, with special guest joining on stage. \ No newline at end of file diff --git a/content/sessions/1113.md b/content/sessions/1113.md deleted file mode 100755 index 8b6ea2cd2..000000000 --- a/content/sessions/1113.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Hatching the Clutch - A Guide to the Apache Incubator" -date: "2019-09-10T12:15:00" -track: "community" -presenters: "Dave Fisher" ---- - -What is the Apache Incubator? What is required from projects in order to graduate? What resources are available and which are required? What does the Foundation provide? Why is the dev email list so important? What about your Mentors? We are all volunteers and strive to be your gateway into taking full advantage of what the Foundation has to offer. Let's review the state of the Clutch and see what is happening now. \ No newline at end of file diff --git a/content/sessions/1114.md b/content/sessions/1114.md deleted file mode 100755 index 8479521da..000000000 --- a/content/sessions/1114.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Taming the Herd - Tuning ATS for linear video" -date: "2019-09-10T14:30:00" -track: "contentdelivery" -presenters: "Evan Zelkowitz" ---- - -At Comcast we have a unique challenge of serving both static and live linear content via our CDN. In general a thundering herd (multiple instantaneous simultaneous requests for the same content) is not a large concern for static files barring a few situations like serving automatic pushed update files. However when serving live content the nature of it means you are almost always generating a herd or on the verge of it due to timing. This talk covers some of the issues we have seen while growing our CDN to serve both live and static content and the solutions we have come to with regards to using Apache Traffic Server as our main caching software. \ No newline at end of file diff --git a/content/sessions/1115.md b/content/sessions/1115.md deleted file mode 100755 index 9807961fc..000000000 --- a/content/sessions/1115.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Does Open Source Participation Get Easier Over Time?" -date: "2019-09-10T17:00:00" -track: "community" -presenters: "Adina Crainiceanu, Myrle Krantz, Ellen Friedman, Dinesh A. Joshi, Holden Karau, Justin Mclean, Joan Touzet" ---- - -What’s it like to be a committed member of an active open source community? What are the awesome parts? What are the challenges? Is open source for you? In this talk, presenters with a wide range of experience at Apache will share their stories and advice on open source participation. From starting out to ongoing involvement, from participating in open source as part of your job, to finding time to participate outside of your job, from finding models and mentors in an online, asynchronous communication environment to being such a mentor, from the worst to the best moments and experiences in open source, join us to hear about more than 30 years of combined experience with Apache projects. \ No newline at end of file diff --git a/content/sessions/1116.md b/content/sessions/1116.md deleted file mode 100755 index 65274dbb8..000000000 --- a/content/sessions/1116.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Widening the Circle: The open source journey from user to contributor" -date: "2019-09-12T17:30:00" -track: "community" -presenters: "Aizhamal Nurmamat kyzy, Riona MacNamara" ---- - -Open source draws its strength from the communities that use and build it. It's their diversity of perspective, skills, and accountability that makes individual projects stronger and builds a richer and solid ecosystem. But while open source is used by the entire world, that broad user community is not yet reflected in the contributor base. In fact, diversity in open source is significantly worse than in proprietary software. While we claim that contribution is open to all, clearly not everybody feels empowered or welcome to contribute to open source projects. And that's a problem. How can we fix this? How can we enable all developers everywhere to contribute to open source and take advantage of the opportunities it presents? In Q1 2019, Google's Open Source Strategy team commissioned a user research study to better understand why users do (or do not) contribute back to the project, and how documentation can help remove roadblocks to that project's adoption and usage - and encourage and enable community contributions. In this talk, we'll cover the methodology and participant profiles of the study, and describe the four essential user personas we identified - along with their critical user-to-contributor journeys. And we'll also distill the research to provide concrete recommendations and best practices for creating documentation that helps your community flourish and your project thrive. \ No newline at end of file diff --git a/content/sessions/1117.md b/content/sessions/1117.md deleted file mode 100755 index 500534d60..000000000 --- a/content/sessions/1117.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Geospatial Data Management in Apache Spark" -date: "2019-09-09T10:00:00" -track: "geospatial" -presenters: "Jia Yu, Mohamed Sarwat" ---- - -The volume of spatial data increases at a staggering rate. This talk comprehensively studies how existing works, such as GeoSpark, extend Apache Spark to uphold massive-scale spatial data. During this talk, we first provide a background introduction of the characteristics of spatial data and the history of distributed data management systems. A follow-up section presents the common approaches used by the practitioners to extend Spark and introduces the vital components in a generic spatial data management system. The third and fourth sections then discuss the ongoing efforts and experience in spatial-temporal data and spatial data analytics, respectively. The fifth part finally concludes this talk to help the audience better grasp the overall content and points out future research directions. \ No newline at end of file diff --git a/content/sessions/1118.md b/content/sessions/1118.md deleted file mode 100755 index 67c2343ad..000000000 --- a/content/sessions/1118.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "The First Mile -- Edge and IoT Data Collection with Apache NiFi and MiNiFi" -date: "2019-09-09T16:30:00" -track: "internetofthings" -presenters: "Andy LoPresto" ---- - -Apache NiFi provided a revolutionary data flow management system with a broad range of integrations with existing data production, consumption, and analysis ecosystems, all covered with robust data delivery and provenance infrastructure. Now learn about the follow-on project which expands the reach of NiFi to the edge, Apache MiNiFi. MiNiFi is a lightweight application which can be deployed on hardware orders of magnitude smaller and less powerful than the existing standard data collection platforms. With both a JVM compatible and native agent, MiNiFi allows data collection in brand new environments — sensors with tiny footprints, distributed systems with intermittent or restricted bandwidth, and even disposable or ephemeral hardware. Not only can this data be prioritized and have some initial analysis performed at the edge, it can be encrypted and secured immediately. Local governance and regulatory policies can be applied across geopolitical boundaries to conform with legal requirements. And all of this configuration can be done from central command & control using an existing NiFi with the trusted and stable UI data flow managers already love. \ No newline at end of file diff --git a/content/sessions/1119.md b/content/sessions/1119.md deleted file mode 100755 index 8c93642c8..000000000 --- a/content/sessions/1119.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Recent advances in Natural Language Processing and Deep Learning" -date: "2019-09-12T17:30:00" -track: "machinelearning" -presenters: "Thomas Delteil" ---- - -In recent years, Machine Learning (ML) approaches to Natural Language Processing (NLP) drastically changed. From a feature-based modelling, where NLP tasks are tackled by using a set of manually engineered features, we moved to the so-called representation learning: text is naturally represented as a sequence of symbols (e.g., words, characters, etc.) and the role of learning how these symbols should be represented is entirely left to the ML method. Deep Learning dominates such paradigmatic shift.nIn this presentation we will review the Deep Learning approach to NLP and discuss different alternatives for representing text, including word embeddings[1] and character embeddings[2].nFinally, we will focus on the latest advances, including: ELMo[3], a model for generating deep contextualized word embeddings, and BERT[4], a very recent Neural Network Model that allows to achieve state-of-the-art results on several NLP tasks with very little task-specific fine-tuning.nThis presentation will mix scientific theory and actual code examples from Apache MXNet (incubating) demonstrating how to leverage these advanced techniques to solve NLP tasks. [1] Word2Vec - Efficient Estimation of Word Representations in Vector Space, Mikolov et al. 2013n[2] CNN character embedding layer - Character-Aware Neural Language Models, Kim et al. 2015n[3] ELMo - Deep contextualized word representations, Peters et al. 2018n[4] BERT - BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding, Devlin et al. 2018 \ No newline at end of file diff --git a/content/sessions/1120.md b/content/sessions/1120.md deleted file mode 100755 index e0c343992..000000000 --- a/content/sessions/1120.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "How to Become an IoT Developer (and Have Fun!)" -date: "2019-09-09T09:00:00" -track: "internetofthings" -presenters: "Justin Mclean" ---- - -I started off my life as a developer writing machine code and C and working on some low-level hardware projects. Then this thing called the internet come along and I moved into the web application space for a couple of decades. More recently I've moved back into commercial IoT development and not unexpectedly a lot has changed over that time. In this talk, I'll cover what it's like developing IoT projects. I'll go over the tools you need and the protocols you need to be familiar with. I'll look at how the C language has evolved to what it is today and how to write code that works well on memory constrained devices. I'll go over producing prototypes, rapid development, debugging and testing embedded applications and what and how much electronics you should learn. In short, everything you need to know in becoming an IoT developer and have fun doing it. \ No newline at end of file diff --git a/content/sessions/1121.md b/content/sessions/1121.md deleted file mode 100755 index dd220bebb..000000000 --- a/content/sessions/1121.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "TVM: An Automated End-to-End Optimizing Compiler for Deep Learning" -date: "2019-09-12T09:00:00" -track: "machinelearning" -presenters: "Tianqi Chen" ---- - -There is an increasing need to bring machine learning to a wide diversity of hardware devices from the datacenter to the edge. Current frameworks rely on vendor-specific operator libraries and optimize for a narrow range of server-class GPUs. Deploying workloads to new platforms such as mobile phones, embedded devices, and accelerators (e.g., FPGAs, ASICs) requires significant manual effort. In this talk, we will talk about Apache TVM -- an end to end optimizing deep compiler stack that brings deep learning models on diverse hardware back-ends that are competitive with state-of-the-art hand-tuned libraries for low-power CPU, mobile GPU, and server-class GPUs. \ No newline at end of file diff --git a/content/sessions/1122.md b/content/sessions/1122.md deleted file mode 100755 index f9a927ff1..000000000 --- a/content/sessions/1122.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache Tomcat / HTTPD Configuration Management" -date: "2019-09-11T15:15:00" -track: "tomcat" -presenters: "Steve Sanders" ---- - -Does your organization have tons of applications and containers? Has this fleet grown to be completely unmanageable? Do undocumented changes cause downtime and pain for your operations? Using modern configuration management tools and practices (Chef and CI/CD), Mastercard created a set of cookbooks that builds, deploys and manages model driven middleware infrastructure. Join us for a talk where we will discuss what is and why you should do configuration management, how we chose to model the middleware configurations, examples of how to test the heck out of the whole assembly and how this methodology brings everything a step closer to “gitops”. This talk will dive into the guts of the cookbook and the processes around how teams onboard to the flow. \ No newline at end of file diff --git a/content/sessions/1123.md b/content/sessions/1123.md deleted file mode 100755 index 92d5ddb34..000000000 --- a/content/sessions/1123.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache NetBeans - Shameless Marketing Tool" -date: "2019-09-12T09:00:00" -track: "community" -presenters: "Anton Epple" ---- - -NetBeans has completed its transition to Apache and is now a top level Apache project with a strong and dedicated community and millions of users worldwide. NetBeans always had great support for Apache Maven & having an IDE of our own is a great chance to further promote Apache projects to a large audience. In this session I'll show you how to plugin your own language, tool, library, server, database or framework and make it easy for developers to get started with them. Use NetBeans as a marketing tool to shamelessly plug your own cool project. \ No newline at end of file diff --git a/content/sessions/1124.md b/content/sessions/1124.md deleted file mode 100755 index 32bc38f89..000000000 --- a/content/sessions/1124.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache Kafka vs. Integration Middleware (MQ, ETL, ESB) - Friends, Enemies or Frenemies?" -date: "2019-09-12T10:00:00" -track: "integration" -presenters: "Kai Waehner" ---- - -MQ, ETL and ESB middleware are often used as integration backbone between legacy applications, modern microservices and cloud services. This introduces several challenges and complexities like point-to-point integration or non-scalable architectures. This session discusses how to build a completely event-driven streaming platform leveraging Apache Kafka’s open source messaging, integration and streaming components to leverage distributed processing, fault-tolerance, rolling upgrades and the ability to reprocess events. nLearn the differences between an event-driven streaming platform leveraging Apache Kafka and middleware like MQ, ETL and ESBs – including best practices and anti-patterns, but also how these concepts and tools complement each other in an enterprise architecture. \ No newline at end of file diff --git a/content/sessions/1125.md b/content/sessions/1125.md deleted file mode 100755 index bd9cc8ddc..000000000 --- a/content/sessions/1125.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Continuous Machine and Deep Learning at Scale with Apache Ignite" -date: "2019-09-12T09:00:00" -track: "machinelearning" -presenters: "Denis Magda" ---- - -With most machine learning (ML) and deep learning (DL) frameworks, it can take hours to move data, and hours to train models. It's also hard to scale, with data sets increasingly being larger than the capacity of any single server. The size of the data also makes it hard to incrementally test and retrain models in near real-time to improve results. Learn how Apache Ignite and GridGain help to address these limitations with model training and execution, and help achieve near-real-time, continuous learning. It will be explained how ML/DL work with Apache Ignite, and how to get started. Topics include:n n— Overview of distributed ML/DL including design, implementation, usage patterns, pros and consn— Overview of Apache Ignite ML/DL, including prebuilt ML/DL, and how to add your own ML/DL algorithmsn— Model execution with Apache Ignite, including how to build models with Apache Spark and deploy them in Igniten— How Apache Ignite and TensorFlow can be used together to build distributed DL model training and execution \ No newline at end of file diff --git a/content/sessions/1126.md b/content/sessions/1126.md deleted file mode 100755 index b7afc2f0b..000000000 --- a/content/sessions/1126.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Day to day with Cassandra: The weirdest and complex situations we found!" -date: "2019-09-11T15:15:00" -track: "cassandra" -presenters: "Carlos Rolo" ---- - -Every Cassandra operator has been hit with a couple of weird/complex cases that don't fit the normal expected failure situation. It can be a problem in hardware, software, networking, operator mistake, or a mix of it all. In this talk we will go through a compilation of such cases that we faced. How do they appear, how did we debug them and how did we fix them. We expect this to be a walk through weird, fun cases and sharing knowledge on the situations and on the fixing of such problems \ No newline at end of file diff --git a/content/sessions/1127.md b/content/sessions/1127.md deleted file mode 100755 index 62886d848..000000000 --- a/content/sessions/1127.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Lessons learned while migrating 3000 openvz containers to CloudStack" -date: "2019-09-10T11:15:00" -track: "cloudstack" -presenters: "Tyler Wilson, Jeffrey Kriegsman" ---- - -When Leaseweb acquired a new company 3000 public cloud customers we had to come up with a strategy to move all these VMs to our CloudStack setup. During this talk we will share: how we migrated a large foreign openvz environment into our kvm cloudstack environment in an automated fashion minimizing customer impact. The lessons we learned along the way and how we would do it again in the future. \ No newline at end of file diff --git a/content/sessions/1128.md b/content/sessions/1128.md deleted file mode 100755 index 6d7ee0a8c..000000000 --- a/content/sessions/1128.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "RocksDB Based Storage Engine Recent Improvements" -date: "2019-09-10T14:30:00" -track: "cassandra" -presenters: "Penchao Wang" ---- - -In last year's NGCC session 'Cassandra Pluggable Storage Engine', Instagram Cassandra team presented a new RocksDB based Cassandra storage engine which brings us huge performance win. Since then our team have been actively working on the new storage engine for adding missing features and further improving efficiency. In this session I will talk about our work on following sections and their impacts to our production system.n* Large disaggregated flash disk supportn* RocksDB based streaming improvementsn* Decoding improvementsn* Space amplification improvementsn* Some missing features added \ No newline at end of file diff --git a/content/sessions/1129.md b/content/sessions/1129.md deleted file mode 100755 index ad67620ee..000000000 --- a/content/sessions/1129.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Observability in distributed system" -date: "2019-09-10T12:15:00" -track: "observability" -presenters: "Willem Jiang, Liang Zhang" ---- - -Observability is quite important for the cloud native application. Itncould be a challenge if we build up the system observer from thenstrache. By leveraging the Skywalking's extension mechanism, we couldnsend out the data to the backend directly to improve thenObservability of the Cloud Native Service. In distributed applications, horizontal sharding and transactio coordination are most popular solutions. Observability for businessninvocation on sharding scenario are not enough; but observer allnsharded details of backend infrastructures indifferently are tooncomplicated. We need a whole picture of SQL and transactions execution status. ShardingSphere is a transport databases middleware for sharding and aggregation service solution; ServiceComb is a micro service and distributed transaction solution. In this talk we will share our integration with Skywalking, which can observe a modern, complicated, distributed system immediately. \ No newline at end of file diff --git a/content/sessions/1130.md b/content/sessions/1130.md deleted file mode 100755 index adb74fb65..000000000 --- a/content/sessions/1130.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Riding Camel in Huawei Cloud" -date: "2019-09-12T16:30:00" -track: "integration" -presenters: "Willem Jiang" ---- - -Apache Camel provides a powerful EIP (Enterprise Integration Patterns) DSL to integrate tons of different middlewares. Now more and more enterprise user moving their application into cloud, it could be a challenge for them to bridge the public cloud service and the inhouse build application from scratch. With the help of Apache Camel, we provide an integration services on Huawei Cloud to address this across cloud integration issues and extend integration ability by leveraging the community collaboration. In this talk, the presenter will share the journey of riding Apache Camel with the integration service of Huawei, and an end to end demo of smart park project. \ No newline at end of file diff --git a/content/sessions/1131.md b/content/sessions/1131.md deleted file mode 100755 index 76220315e..000000000 --- a/content/sessions/1131.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Using PLC4X and LF Edge EVE to build a next generation Industry 4.0 IoT Platform" -date: "2019-09-09T14:00:00" -track: "internetofthings" -presenters: "Roman Shaposhnik, Christofer Dutz" ---- - -Industrial IoT platforms are undergoing rapid shift away from vertically integrated systems built mostly out of embedded software towards an Edge Virtualization Platforms and commodity, cloud native software. While the cloud, big-data, fast-data, machine-learning has been a huge thing in the IT industry for the last years, the production industry was stuck in a proprietary world, dominated by some of the big players. Two open source projects now offer the building blocks for the new platform that can hope to break that proprietary chokehold. Linux Foundation's Project EVE (Edge Virtualization Engine) aims to develop an open and standardized edge container runtime platform capable of orchestrating cloud-native applications across the enterprise, on-premises edge and IoT deployments. It does so by leveraging hardware-assisted virtualization and provides software defined networking and I/O virtualization capabilities to the end user applications running on it. We use EVE as a runtime for the new Apache PLC4X project that in turn provides a unified API for communicating with industrial controllers via a wide variety of protocols. Besides providing the APIs and driver implementations, PLC4X also provides integration modules to other Apache frameworks such as Apache Edgent, Apache Camel and Apache Kafka that are also supported by the unified EVE IoT edge container runtime. This makes it really easy to integrate industrial controllers into our Open-Source world. Self-optimizing production processes, predictive maintenance, industrial control systems running in the companies cloud. This was all almost impossible until now. With Apache PLC4X and LF Edge EVE it will be easy. \ No newline at end of file diff --git a/content/sessions/1132.md b/content/sessions/1132.md deleted file mode 100755 index 319675598..000000000 --- a/content/sessions/1132.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Schema-Controlled HDFS Column Encryption and Use Cases" -date: "2019-09-11T12:00:00" -track: "bigdata" -presenters: "Mohammad Islam, Xinli Shang" ---- - -Motivation: Access control via encryption improves security coverage compared with traditional enforcement in the access path because encryption can prevent invalid accesses from any angle. Finer-grained access control at the column-level is needed because in a typical big dataset, only a few columns are sensitive and need to be protected, and different columns could have different sensitivities and a different set of eligible readers. Design: With encryption features in columnar file format like Apache Parquet, column access control via encryption becomes possible. But to adopt these features into existing analytic frameworks that might use Apache Hive, Apache Spark, Apache Preto etc, is a challenge because significant changes are needed to those query engines to control the encryption. n nTo avoid massive changes in existing frameworks, the Apache Parquet community designed a schema controlled column encryption mechanism. The schema of data tables can be leveraged by a system architect in order to define the sensitivity of a column, that will be propagated through the stack and will eventually trigger the encryption on that column in the Apache Parquet writing layer. This solution is applicable in many analytic frameworks via transparent plug-in invocation that avoids massive changes in the frameworks, making Apache Parquet encryption easy to adopt. This mechanism is possible to be extended to support Apache ORC too. n nUse Cases: 1. HDFS ingesting pipelines with Spark and Hudi encrypt sensitive columns automatically with schema-controlled crypto settings. n nIn this use case, we will talk about how to use schema to control column encryption in Apache Parquet, and what is the flow that the pipeline can automatically encrypt the columns once schema sets the sensitivity. n n2. Column access control scalability and performance analysis for analytic frameworks with Apache Hive and Presto on Apache Parquet.n nLike Apache Spark, Apache Hive and Presto are two popular query engines. We will show our analysis on scalability and performance of column access control in an analytic framework. n nSummary: In this talk, we will present the motivation and design of schema-controlled column encryption at file format level, and how it can ease the adoption of encryption features to current popular query engines like Apache Hive, Apache Spark and Presto. Use cases will be presented to show how to use schema controlled column encryption in analytic pipeline and frameworks. \ No newline at end of file diff --git a/content/sessions/1133.md b/content/sessions/1133.md deleted file mode 100755 index 69dcde466..000000000 --- a/content/sessions/1133.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "From an idea to an Apache TLP" -date: "2019-09-12T14:00:00" -track: "community" -presenters: "Christofer Dutz" ---- - -About 3 years ago, I had an idea of using Open-Source software to create the next generation of industrial it solutions. At ApacheCon 2017 in Miami I introduced this idea to the public with my 'Building SCADA systems with Apache Software'. 2019 the Apache PLC4X project is an Apache TLP. In this talk I will not talk about technical details of the project itself, but all the steps we took on this journey from a community-building point of view. Starting way before writing the first line of code. After these last two years, I would claim that 'community-building' is by far the most challenging task when initiating a new project, but it's also seems to be one we tend to treat as second-class citizens - even in established projects. Hopefully I will be able to show you how rewarding community-building can be ... After all: Apache is all about community over code and it's that way for a reason \ No newline at end of file diff --git a/content/sessions/1134.md b/content/sessions/1134.md deleted file mode 100755 index b068a1bca..000000000 --- a/content/sessions/1134.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Ceph with CloudStack" -date: "2019-09-10T12:15:00" -track: "cloudstack" -presenters: "Andrija Panić" ---- - -In this presentation, we are going to give a brief introduction to Ceph and cover some considerations around it's architecture, both in general and related to CloudStack. We are going to cover Ceph's integration into CloudStack, compare it's feature set versus other Primary Storage solutions for CloudStack and also share some general advises on it's setup. \ No newline at end of file diff --git a/content/sessions/1135.md b/content/sessions/1135.md deleted file mode 100755 index eccbf2a3e..000000000 --- a/content/sessions/1135.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "HTTP/2, HTTP/3 and SSL/TLS State of Art in Our Servers" -date: "2019-09-11T14:15:00" -track: "tomcat" -presenters: "Jean-Frederic Clere" ---- - -A new protocol is getting ready HTTP/3 we will look to where we are with it in our serves.nThe 'old' HTTP/2 protocol and the corresponding TLS/SSL are common to Traffic Server, HTTP Server and Tomcat.nThe presentation will shortly explain the new protocol and the ALPN extensions and look to the state of the those in our 3 servers and show the common parts and the specifics of each servers.nA demo configuration of each server will be run. \ No newline at end of file diff --git a/content/sessions/1136.md b/content/sessions/1136.md deleted file mode 100755 index b1e5eab59..000000000 --- a/content/sessions/1136.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache httpd and TLS/SSL certificates validation" -date: "2019-09-12T17:30:00" -track: "general" -presenters: "Jean-Frederic Clere" ---- - -We will look to 2 different things here, validation of the server certificate andnvalidation of the client certificates. For the server certificate we will show Let's encrypt and mod_md and speak aboutnthe new ACMEv2 protocol and OCSP stapling. For the client certificates we look to OCSP and other validations. Demo and quick start example will provided during the talk. \ No newline at end of file diff --git a/content/sessions/1137.md b/content/sessions/1137.md deleted file mode 100755 index aaabd347a..000000000 --- a/content/sessions/1137.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Backup and Recovery in CloudStack (it's Here!)" -date: "2019-09-10T12:15:00" -track: "cloudstack" -presenters: "Paul Angus" ---- - -This presentation will demonstrate the long awaited Backup & Recovery Framework feature. The presentation will feature a real-world demonstration as well explain the design philosophy and how the feature has been built to be agnostic of the backend Backup and Recovery software in use. \ No newline at end of file diff --git a/content/sessions/1138.md b/content/sessions/1138.md deleted file mode 100755 index 77a1aae58..000000000 --- a/content/sessions/1138.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Building Data Platform for your Next Meetup Event with Apache Foundation on Cloud" -date: "2019-09-11T17:45:00" -track: "bigdata" -presenters: "Chengzhi Zhao" ---- - -One of the challenges at Meetup is how to build a scalable, reliable and efficient data platform to help our ML team builds models that recommend events fit your interests. With emerging sophisticated batch and streaming frameworks and cloud solutions, our data platform went through massive changes in the past two years. In this talk, I'll discuss the evolution of how Meetup data platform utilizes Apache-based data systems, including Sqoop, Hive, Flume, Spark, Flink, Beam, Airflow. I'll talk about architecture changes to our batch and stream pipeline solutions and what pros/cons to move data platform 100% to cloud. I'll also share some lessons we learned and best practices on building distributed systems for data platform, and how the data platform collaborates with machines learning and data science team. \ No newline at end of file diff --git a/content/sessions/1139.md b/content/sessions/1139.md deleted file mode 100755 index cae1a3a53..000000000 --- a/content/sessions/1139.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Transaction Management on Cassandra" -date: "2019-09-10T17:00:00" -track: "cassandra" -presenters: "Hiroyuki Yamada, Jay Zhuang, Dikang Gu" ---- - -Scalar DB is an open source library released under Apache 2 which realizes ACID-compliant transactions on Cassandra, without requiring any modifications to Cassandra itself. It achieves strongly-consistent, linearly scalable, and highly available transactions. This talk will present the theory and practice behind Scalar DB, as well as providing some benchmark results and use cases. \ No newline at end of file diff --git a/content/sessions/1140.md b/content/sessions/1140.md deleted file mode 100755 index 83a1a7b4b..000000000 --- a/content/sessions/1140.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache OFBiz in the Cloud(s)" -date: "2019-09-12T09:00:00" -track: "ofbiz" -presenters: "Deepak Dixit" ---- - -Cloud computing greatly simplifies provisioning, scaling, reliability and maintenance: these aspects are particularly important for Enterprise applications like Apache OFBiz. This talk will provide effective tips, based on experience and lessons learned in the field, to enable OFBiz for the cloud. Topics covered include OFBiz specific solutions for:n- deployments in AWSn- images - AMI, Dockern- continuous deploymentsn- configurations for load balancersn- session management - sticky sessions, distributed sessions (with Redis)n- database options for the cloud - RDS (AWS), Google Cloud Spanner \ No newline at end of file diff --git a/content/sessions/1141.md b/content/sessions/1141.md deleted file mode 100755 index 011b6b41e..000000000 --- a/content/sessions/1141.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "State of integration with Apache Camel" -date: "2019-09-11T15:15:00" -track: "integration" -presenters: "Claus Ibsen" ---- - -Apache Camel is the leading open source integration framework, which has been around for over a decade. In this talk we will look back in history, to understand how the integration landscape has evolved from EAI, SOA, and ESB architectures up to microservices, and now with modern serverless and cloud native platforms. Apache Camel has been along for the ride. And we will look to the future and see how the latest release v3 of Apache Camel, is aimed for running modern cloud native workloads with Camel K. In this talk you will:n- Learn from history software integration, and why you should rely on existing, proven fully featured integration frameworks instead of rolling out your own DIY solutionsn- See how software integration is (still) important in today’s modern architectures and what role does Camel have in the new cloud native worldn- What is new and noteworthy in Apache Camel version 3 \ No newline at end of file diff --git a/content/sessions/1142.md b/content/sessions/1142.md deleted file mode 100755 index 67decc0f7..000000000 --- a/content/sessions/1142.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache Camel K: a cloud-native integration platform" -date: "2019-09-11T17:45:00" -track: "integration" -presenters: "Nicola Ferraro, Andrea Tarocchi" ---- - -In this session we are going to introduce the latest innovation from the Apache Camel community: Camel K, a lightweight integration platform, born on Kubernetes, with serverless superpowers.nCamel K enables developers that want to integrate systems to directly write Camel DSL code in the cloud, with a great developer experience and really fast turnaround times.nYou’ll see Camel K in action with a live coding demo that will explore the main features that it provides.nYou’ll also learn how Camel K works under the hood and will have a glimpse on the future roadmap. \ No newline at end of file diff --git a/content/sessions/1143.md b/content/sessions/1143.md deleted file mode 100755 index 33564a34a..000000000 --- a/content/sessions/1143.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Modern UI for CloudStack" -date: "2019-09-09T10:00:00" -track: "cloudstack" -presenters: "Anurag Awasthi, Rohit Yadav" ---- - -The CloudStack UI code has tolerated massive technical debt over the years and the UI codebase has reached a breaking point. In this talk, we will do a comparison of modern UI development techniques, the pros and cons of current UI UX and codebase, ideas and experiments of a new UI for CloudStack. We need to discuss and propose a new UI that:n* Has CloudMonkey like api discovery and show/hide/block, buttons/sections based on what is allowedn* URL router to navigate to resources, and clicking back button does not close the UIn* Easy to maintain, implement and extend via plugins etc.n* Does not take months to develop and port existing implementation to the new onen* Configurable, theme-able, don't use our own UI framework, but consider something like Vue/React/Angular etc.n* Explore a components based idea, show an api-auto-discovery approach to auto-generate boiler plate for UI component based on an API's request and response paramsn* Make UI more communicative about underlying infra state in real time with modern graphical libraries \ No newline at end of file diff --git a/content/sessions/1144.md b/content/sessions/1144.md deleted file mode 100755 index dd84e6644..000000000 --- a/content/sessions/1144.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "CloudStack identity federation: An overview" -date: "2019-09-10T15:30:00" -track: "cloudstack" -presenters: "Rafael Weingärtner" ---- - -Federated authentication in online services is increasingly being used, whether through public identity providers such as Facebook, Google and others, or in organizations' private identity providers. Identity federation is a technology that has come to simplify the life of the user, but that does not always simplify the life of application developers and operators. The purpose of this talk is to share a bit the challenges, benefits and problems of using federated authentication with multiple identity providers in Apache Cloudstack. \ No newline at end of file diff --git a/content/sessions/1145.md b/content/sessions/1145.md deleted file mode 100755 index dfef0ca1f..000000000 --- a/content/sessions/1145.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Serving Clean URLs via Custom Internet Media Types on Tomcat" -date: "2019-09-10T15:30:00" -track: "tomcat" -presenters: "Garret Wilson" ---- - -Increasingly static site generators are being used to create a presence on the web, both for the speed and simplicity of the resulting static pages. A related trend is the use of 'clean' URLs with no resource name extensions, such as `/products/mousetrap` instead of `/products/mousetrap.html`. Clean URLs are more memorable, less fragile, do not leak implementation information, and are more search engine friendly. Because the default Tomcat implementation determines MIME type by filename extension, 'clean' filenames are traditionally produced by servlet mapping or URL rewriting, perhaps coupled with Apache HTTP Server in front of Tomcat using `mod_rewrite`. This presentation shows how Tomcat has been extended to dynamically determine MIME type at runtime based on metadata stored in a parallel file tree or as sidecar files, and to serve the pages with clean resource names directly from Tomcat with no need for URL rewriting. A tutorial on the static site generator Guise Mummy is also included, illustrating how to generate a static site with clean resource names, host the site immediately using embedded Tomcat, and/or deploy the site to AWS S3 using bucket object metadata with a single command. \ No newline at end of file diff --git a/content/sessions/1146.md b/content/sessions/1146.md deleted file mode 100755 index 1c02f12f9..000000000 --- a/content/sessions/1146.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Building an open source analytics distribution on Kubernetes that meets the needs of a modern data scientist." -date: "2019-09-12T17:30:00" -track: "machinelearning" -presenters: "Jay Vyas" ---- - -Modern machine learning libraries such as Jupyter, Kubeflow, PyTorch on technologies that are managed outside of the linux foundation, often forcing users to pay for cloud or hosted offerings to manage integration and tool dependencys. This challenges the ability of the ASF to continue blazing a trail of open innovation in the data science and analytics frontiers which are transforming so many businesses today. New projects in the ASF: such as Airflow, PredictionIO, SystemML, and NiFi - combined with technologies such as Spark and HDFS can be used to build an end-to-end analytics solution that is flexible and modular enough to meet the needs of modern data scientists, at pedabyte scale. In this talk, we'll introduce a prototypical 'next generation' prototype which we propose as the future direction of ApacheBigtop, which runs entirely on Kubernetes, to minimize cost and maximize the portability of your modern hybrid analytics architecture. \ No newline at end of file diff --git a/content/sessions/1147.md b/content/sessions/1147.md deleted file mode 100755 index 816792476..000000000 --- a/content/sessions/1147.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "From Postgres to an In-Memory Grid with Apache Ignite" -date: "2019-09-11T16:45:00" -track: "bigdata" -presenters: "Vitor Wakim" ---- - -This talk describes how a food tech company with 50k merchants was at a critical situation by delivering events to devices by using events polling on a Postgres database, and how it could solve this problem by moving the events repository to Apache Ignite using distributed, in-memory SQL. The change also allowed the company to be on the right track to support more than 500k simultaneous merchants. \ No newline at end of file diff --git a/content/sessions/1148.md b/content/sessions/1148.md deleted file mode 100755 index be368b3b6..000000000 --- a/content/sessions/1148.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "I Love Lucee" -date: "2019-09-11T16:45:00" -track: "tomcat" -presenters: "Igal Sapir" ---- - -Lucee is a free and open source application server that implements the CFML language, which was popularized by the commercial product ColdFusion. It allows for rapid application development, and is implemented as a JSP Servlet so it is deployed inside a Servlet Container like Apache Tomcat. The Lucee website: https://lucee.org/ GitHub Repo:nhttps://github.com/lucee/Lucee Support Forum: https://dev.lucee.org/ \ No newline at end of file diff --git a/content/sessions/1149.md b/content/sessions/1149.md deleted file mode 100755 index 22c426915..000000000 --- a/content/sessions/1149.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Data Movement & Integration at PayPal & LinkedIn using Apache Gobblin" -date: "2019-09-12T17:30:00" -track: "integration" -presenters: "Jay Sen, Sudarshan Vasudevan" ---- - -Data replication at PayPal drives various different business use-cases from fraud detection, user behavioral analysis, credit checks to lot of other offline business decisions. During this talk, we will present how Apache Gobblin empowers data movement and integrations at PayPal in partnership with LinkedIn to showcase all the recent features as well as the planned roadmap for the platform. Apache Gobblin is a distributed data integration framework that simplifies common aspects of big data integration such as data ingestion, replication, organization and lifecycle management for both streaming and batch data ecosystems. In the second half of this presentation, we will present recent additions to Gobblin including: 1. A new declarative approach for defining data pipelines using Gobblin-as-a-Service, and 2. Real world experiences running hybrid batch and streaming pipelines using Gobblin. \ No newline at end of file diff --git a/content/sessions/1150.md b/content/sessions/1150.md deleted file mode 100755 index fbc497493..000000000 --- a/content/sessions/1150.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Geospatial Data and Processing - Reusable Building Blocks" -date: "2019-09-09T09:00:00" -track: "geospatial" -presenters: "George Percivall" ---- - -Reuse of common elements for geospatial information and processing results in increased productivity, lower interoperability friction, and higher data quality. This presentation provides a survey of reusable geospatial building blocks. Common practices for coordinate reference systems (CRSs), spatial geometries and data arrays used for projects with geospatial content will be described based on open source projects and open standards. Emphasis is placed on the use of open standards including the recently updated OGC CRS Well Known Text (CRS WKT) and OGC APIs. The presentation will provide the latest update on OGC API development. OGC APIs are being defined for geospatial resources, e.g., maps, features, coverages. Developed using OpenAPI, the APIS can be implemented in a number of languages and patterns. The presentation will be describe the state of implementations and plans for standardization. The modular structure enables flexibility for developers to reuse OGC APIs in their APIs. This presentation will allow time for discussion of coordination across Apache projects with geospatial content. Geospatial tracks at previous ApacheCons concluded with an open discussion leading to the creation of geospatial@apache.org. \ No newline at end of file diff --git a/content/sessions/1151.md b/content/sessions/1151.md deleted file mode 100755 index adc3105d4..000000000 --- a/content/sessions/1151.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache When You’re Not In Charge" -date: "2019-09-11T12:00:00" -track: "community" -presenters: "Bob Paulin" ---- - -Apache has a 20 year history of building resilient and sustainable communities of software developers through the Apache Way. While some companies are adopting Inner sourcing, other companies still have a taboo surrounding open source culture. Many of us work at these open source agnostic companies and don’t have manager in our title. Are those of us in those companies doomed to a life of Dilbert? You don’t have to be! This talk is all about how to bring a little Apache culture to your Apache oblivious team. This talk will cover how a little Apache Way goes a long way solving common team problems with knowledge sharing, technical/personal disputes, mentorship, and shipping software. So even though you’re the only one on your team that “gets” Apache, everyone can share in the rewards of being a part of an effective team. \ No newline at end of file diff --git a/content/sessions/1152.md b/content/sessions/1152.md deleted file mode 100755 index d48724e80..000000000 --- a/content/sessions/1152.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Going Beyond StringUtils: A Broader Look at Apache Commons" -date: "2019-09-09T15:00:00" -track: "general" -presenters: "Rob Tompkins" ---- - -The Apache Commons Project is home to one of the more widely consumed Java utility functions, namely “StringUtils.isBlank(String str);”. But when you look beyond commons-lang, you’ll find a wide array of heavily utilized and extremely useful tools. For example, commons-cli is the command line parser used by the Apache Maven project; commons-fileupload is heavily used by both Apache Tomcat and Apache Struts; commons-text is heavily used by Apache Hadoop; and commons-dbcp is a database connection pool used across Tomcat as well as the Spring Framework. We’ll look at some common problems facing developers today and how the Apache Commons Project can address them, diving into specific examples as we go. \ No newline at end of file diff --git a/content/sessions/1153.md b/content/sessions/1153.md deleted file mode 100755 index 82f9822d6..000000000 --- a/content/sessions/1153.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Life beyond Kafka with Apache Pulsar" -date: "2019-09-10T11:15:00" -track: "internetofthings" -presenters: "Alvaro Santos Andrés, Javier Santos Andres" ---- - -In a world where everything is connected, trillions of devices are sending data in real-time to data centers. Apache Kafka as a streaming platform is the system by excellence for many years. However, there are many competitors who are making its life more difficult. In this talk, I will introduce to one of them, Apache Pulsar sponsored by Yahoo itself. Pulsar is the next generation messaging system that uses a different architecture to achieve durability, performance, scalability, efficiency, multi-tenancy and geographic replication. Besides, I will focus on why Apache Pulsar is a great streaming platform in IoT projects. \ No newline at end of file diff --git a/content/sessions/1154.md b/content/sessions/1154.md deleted file mode 100755 index 2da635f18..000000000 --- a/content/sessions/1154.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache Commons State of the Union" -date: "2019-09-09T14:00:00" -track: "general" -presenters: "Rob Tompkins" ---- - -We will go over the state of the Apache Commons Project. Where things stand, and will take a look at the future road map for the project. \ No newline at end of file diff --git a/content/sessions/1155.md b/content/sessions/1155.md deleted file mode 100755 index a8470db5b..000000000 --- a/content/sessions/1155.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Exploring github data using Apache Drill running on ARM64" -date: "2019-09-11T17:45:00" -track: "drill" -presenters: "Ganesh Raju" ---- - -Inspired by Google's Felipe Hoffa's effort/talk, this attempt is to replicate an usecase of exploring github data using Apache Drill and lessons learnt. Apache Drill is a distributed MPP query layer that supports SQL and alternative query languages against NoSQL and Hadoop data storage systems. The talk will highlight steps to build Apache Drill from source code, and showcase a demo analyzing terabytes of github data (issues and comments), running on a cluster of 3 node ARM64 servers. \ No newline at end of file diff --git a/content/sessions/1156.md b/content/sessions/1156.md deleted file mode 100755 index 936d3210c..000000000 --- a/content/sessions/1156.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache httpd reverse proxy and Tomcat:" -date: "2019-09-11T17:45:00" -track: "tomcat" -presenters: "Jean-Frederic Clere" ---- - -Although mostly known as a fast and reliable web server, Apache httpd also excels as a reverse proxy.nIn this session find out how to setup httpd as a reverse proxy, how to connect to Tomcat using HTTP and AJP.nWe will also look to the full feature list of Apache httpd proxying capability. \ No newline at end of file diff --git a/content/sessions/1157.md b/content/sessions/1157.md deleted file mode 100755 index 39a5905aa..000000000 --- a/content/sessions/1157.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "The future of the CloudStack Virtual Router" -date: "2019-09-10T11:15:00" -track: "cloudstack" -presenters: "Rohit Yadav" ---- - -This talk surveys and explores the scopes and design for the next-generation network model consolidation in CloudStack and a new super-fast micro-vm based VR. The talk will discuss the scope and goals for the next-generation VR, consolidation of basic and advanced networks and zones, flexible networking topologies, UI based network designing, a standard systemvm patching mechanism and uniform VR programming interface with a lightweight secured agent, faster rules programming and zero-downtime upgrades. It will also explore new abilities to allow live/online patching, containerization of VR processes and pluggability of user-imported data/router appliances (bring your own X) such as pfsense, vyos etc for VPN, FTP, DNS, DHCP, etc. \ No newline at end of file diff --git a/content/sessions/1158.md b/content/sessions/1158.md deleted file mode 100755 index eada2c95e..000000000 --- a/content/sessions/1158.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "What's New and Coming to Apache Hivemall: Building More Flexible Machine Learning Solution for Apache Hive and Spark" -date: "2019-09-12T16:30:00" -track: "machinelearning" -presenters: "Takuya Kitazawa, Makoto Yui" ---- - -Apache Hivemall is a scalable machine learning library for Apache Hive, Spark, and Pig. Hivemall allows us to apply a wealth of machine learning techniques to massive data stored in distributed storage by just writing a series of SQL-like queries. It provides classification, regression, recommendation, anomaly detection, and topic modeling functionalities in a scalable manner, along with a variety of auxiliary functions for data preprocessing and feature engineering. This talk demonstrates the Hivemall library with a special emphasis on its new features merged after the first Apache Incubator release. Hivemall v0.5.2-incubating, the latest version as of April 2019, has introduced a state-of-the-art generalized factor model named Field-Aware Factorization Machines and many useful UDFs (e.g., data sketching) originated from the Brickhouse Hive UDF package. We also show the roadmap of this incubating project. Open issues and pull requests include Apache Spark 2.4 support, implementation of new algorithms such as word2vec and multi-nominal logistic regression, as well as integration with widely-used tools like XGBoost and LightGBM. \ No newline at end of file diff --git a/content/sessions/1159.md b/content/sessions/1159.md deleted file mode 100755 index 611b9456f..000000000 --- a/content/sessions/1159.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Deploying a Java-based Tomcat application to Azure" -date: "2019-09-10T14:30:00" -track: "tomcat" -presenters: "Devlin Duldulao" ---- - -In this talk, you will learn how you can use Azure Pipelines to deploy a Java web application to Apache Tomcat with a MySQL database on Azure. \ No newline at end of file diff --git a/content/sessions/1160.md b/content/sessions/1160.md deleted file mode 100755 index 7193d8656..000000000 --- a/content/sessions/1160.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Unleashing Apache Kafka and TensorFlow in Hybrid Cloud Architectures" -date: "2019-09-10T17:00:00" -track: "machinelearning" -presenters: "Kai Waehner" ---- - -How can you leverage the flexibility and extreme scale in the public cloud combined with your Apache Kafka ecosystem to build scalable, mission-critical machine learning infrastructures, which span multiple public clouds or bridge your on-premise data centre to cloud? nThis talk will discuss and demo how you can leverage machine learning technologies such as TensorFlow with your Kafka deployments in public cloud to build a scalable, mission-critical machine learning infrastructure for data ingestion and processing, and model training, deployment and monitoring.nThe discussed architecture includes capabilities like scalable data preprocessing for training and predictions, combination of different Deep Learning frameworks, data replication between data centres, intelligent real time microservices running on Kubernetes, and local deployment of analytic models for offline predictions. \ No newline at end of file diff --git a/content/sessions/1161.md b/content/sessions/1161.md deleted file mode 100755 index bf49c22ba..000000000 --- a/content/sessions/1161.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Customized CloudStack in Leaseweb" -date: "2019-09-09T10:00:00" -track: "cloudstack" -presenters: "Rakesh Venkatesh" ---- - -In this talk we will share how Leaseweb customized the CloudStack project to serve our business model and our customers, by doing some improvements and do some bug fixes. We always focused on upgrading our fork to the latest in the community while making sure all our improvements are ported to the latest version. In this talk we will tell you about some features and fixes we did like: - multiple NICs feature, - NoVNC, autoscaling, - SSL offloading in load balancer - and other changes made to serve some of our customers' needs. We also love to give the community as we take from it, so we open PRs to release our improvements to the community and based on the community reactions to this talk will start to open new PRs of these changes. \ No newline at end of file diff --git a/content/sessions/1162.md b/content/sessions/1162.md deleted file mode 100755 index 06cea877a..000000000 --- a/content/sessions/1162.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Customer Cloud Portal" -date: "2019-09-09T16:30:00" -track: "cloudstack" -presenters: "Sebastian Bretschneider, Christian Elmer" ---- - -A technical overview showing how we at itelligence as a cloud service provider use CloudStack and how we extent our infrastructure with additional service components to enhance our cloud service portfolio. \ No newline at end of file diff --git a/content/sessions/1163.md b/content/sessions/1163.md deleted file mode 100755 index 8086752d3..000000000 --- a/content/sessions/1163.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache Science Data Analytics Platform (SDAP)" -date: "2019-09-09T11:30:00" -track: "streaming" -presenters: "Thomas Huang" ---- - -An Analytics Center Framework (ACF) is an environment that enables the confluence of resources for scientific investigation. It harmonizes data, tools and computational resources which subsequently enable the research community to focus on the investigation. The Earth science community is an innovative community. We produce many tools and solutions to improve how we do science. In computer science, a framework is a reusable, semi-complete application that can be specialized to produce custom applications [Johnson:88]. After more than two years of actively developing an open source ACF, on October 2017, the NASA AIST OceanWorks project established collaboration with the Apache Software Foundation, called the Apache Science Data Analytics Platform (SDAP). It is a big data analytics platform designed for cloud-based data management, analytics, match-up, and data discovery services. It is a community-support, extensible open source GIS platform. The motivation is to empower the Earth and Space Science Informatics community to develop a common big data solution for the cloud and on-premise cluster. The big data analytics platform is being used to support NASA Sea Level research, GRACE and GRACE Follow-On mission sciences, and NASA Physical Oceanography, etc. This talk describes the Apache SDAP and lesson learned from developing and moving SDAP in production to support various NASA and JPL researches. \ No newline at end of file diff --git a/content/sessions/1164.md b/content/sessions/1164.md deleted file mode 100755 index 5d78dde05..000000000 --- a/content/sessions/1164.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "GeoMesa on top of Accumulo, HBase, Cassandra, and big data file formats for massive geospatial data - a LocationTech Project" -date: "2019-09-09T14:00:00" -track: "streaming" -presenters: "James Hughes, Eddie Pickle" ---- - -LocationTech is the geospatial software working group of the Eclipse Foundation. The projects range from fundamental libraries that provide spatial operations to complex library suites which coordinate multiple Apache projects to build complete spatial processing solutions.nLocationTech GeoMesa builds on top of distributed Apache databases like Accumulo, HBase, Cassandra and Kafka to provide indexing, querying, and analysis for large spatio-temporal datasets. GeoMesa does this by integrating other LocationTech projects like JTS, Spatial4J, and SFCurve with these databases as well as Apache open source file formats such as Avro, Arrow, Orc, and Parquet.nIn this talk, we will give an overview of the geospatial capabilities that the foundational LocationTech libraries can bring to a project. With that background, we will discuss how GeoMesa integrates those capabilities into distributed databases and the file formats. We will wrap up with a quick look at the other big geo-data projects in LocationTech (GeoTrellis, GeoWave, and RasterFrames). \ No newline at end of file diff --git a/content/sessions/1165.md b/content/sessions/1165.md deleted file mode 100755 index 5ab8b1c93..000000000 --- a/content/sessions/1165.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Realtime Geospatial Analytics with GPUs, RAPIDS, and Apache Arrow" -date: "2019-09-09T17:30:00" -track: "streaming" -presenters: "Josh Patterson" ---- - -Prior to RAPIDS, geospatial analytics, especially networking analytics for routing, required large CPU clusters to process. Even with 100s of machines, it would take hours if not days to get answers. With RAPIDS, and Apache 2.0 open source project built on Apache Arrow, graph analytics, clustering, and many other geospatial workflows can be completed end to end in seconds. Learn how to load data (CSV, Parquet, or ORC) directly into GPU memory with cuIO, process it with Dask-cuDF, and analyze it with cuML and cuGraph in seconds, all on a single node. Finally, learn how RAPIDS is scaling to multiple GPU nodes to solve the largest of geospatial challenges. \ No newline at end of file diff --git a/content/sessions/1166.md b/content/sessions/1166.md deleted file mode 100755 index 60c385f94..000000000 --- a/content/sessions/1166.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache Iceberg: a table format for distributed databases" -date: "2019-09-10T11:15:00" -track: "bigdata" -presenters: "Ryan Blue" ---- - -'This talk will introduce Apache Iceberg (Incubating), a new table format designed for distributed databases that can be used directly by any size application. It will cover Iceberg’s design and the features that distinguish it from other table formats, including: -Reliable schema evolutio -Hidden partitioningn-Atomic operationsn-Support for object stores like S3n-Support for direct client access This talk will also present how Iceberg performs at scale in Netflix’s environment.' \ No newline at end of file diff --git a/content/sessions/1167.md b/content/sessions/1167.md deleted file mode 100755 index 31c217a42..000000000 --- a/content/sessions/1167.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Optimizing Big Data Pipelines with Apache Nemo (Incubating)" -date: "2019-09-10T12:15:00" -track: "bigdata" -presenters: "John Youngseok Yang" ---- - -Optimizing scheduling and communication behaviors of big data pipelines for resource and data characteristics is crucial for achieving high performance. For example, optimizing for geographically-distributed resources, cheap transient resources, disk-based large data shuffle, and skewed data have recently received a lot of interest and attention. The new incubating Apache Nemo project aims to make it easy to express and enforce such optimizations by providing a policy interface that transforms an intermediate representation (IR) of data processing applications. Apache Nemo executes the optimized IR on distributed computers while enforcing the specified scheduling and communication behaviors, and at the same time maintaining the correct application semantics. Apache Nemo is closely integrated with existing Apache Big Data projects. Apache Nemo currently supports optimizing Apache Beam and Apache Spark applications, and uses Apache REEF to run on Apache Hadoop YARN and Apache Mesos. In this talk I will describe how to develop and compose new optimization policies on Apache Nemo, how the optimizations are actually enforced in the distributed execution, and the roadmap to further improve Apache Nemo. \ No newline at end of file diff --git a/content/sessions/1168.md b/content/sessions/1168.md deleted file mode 100755 index 3b666bd5c..000000000 --- a/content/sessions/1168.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache Hudi (Incubating) : the past, present and future of efficient data lake architectures" -date: "2019-09-10T14:30:00" -track: "bigdata" -presenters: "vinoth chandar, Balaji Varadarajan" ---- - -Apache Hudi is a newly incubating project at the ASF. Originally created at Uber, to power it’s vast big data lake, Hudi provides key features like atomic writes, snapshot isolation, incremental views, rollbacks, point-in-time restores, file size management & many more. To date, the big data community has been polarized amongst batch and streaming systems, when balancing data freshness and scale. Hudi also addresses a combined need for speed and scale, that does not naturally fit into existing batch and streaming data processing architectures, by way of ability to support continuous ingestion and asynchronous compaction of row & columnar data. In this talk, we will briefly discuss the history of the project - the motivating use-cases & the architectural underpinnings, that spurred the need for such a system. We will examine a blueprint for reliable and state-of-the-art data lake architecture and explain how Hudi’s current capabilities play a central role in it. We will also share hands-on recipes to leverage Hudi in your organization, for popular tasks like data ingestion or ETLs. We will dedicate the remainder of the talk to discussing the roadmap ahead, touching upon areas like external indexing, storage layout organization, as well as advanced topics like building realtime ML feature stores or composing safe multi-stream ETL joins. \ No newline at end of file diff --git a/content/sessions/1169.md b/content/sessions/1169.md deleted file mode 100755 index fd3aea3a0..000000000 --- a/content/sessions/1169.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache Pinot (incubating): Building Realtime Analytics Applications at LinkedIn Scale" -date: "2019-09-10T15:30:00" -track: "bigdata" -presenters: "Jackie Jiang, Seunghyun Lee" ---- - -LinkedIn is the most advantageous social networking tool available to job seekers and business professionals today, with 610+ million members creating millions of posts, videos, and articles that generate tens of millions of shares, comments and likes per day. LinkedIn has leveraged this activity data to build rich interactive user-facing analytics applications like “Who Viewed My Profile”, Talent Insights, Ad Analytics and Publisher Analytics, among others. These applications are all powered by Pinot, as are internal dashboards, anomaly detection and root cause analysis platform like ThirdEye. Apache Pinot (incubating) is a distributed columnar storage engine that can ingest ndata in realtime and serve analytical queries at low latency. In this talk, we will discuss how Pinot has become the de-facto solution for serving analytic queries in milliseconds, ad-hoc reporting, monitoring & Anomaly Detection on multidimensional data. \ No newline at end of file diff --git a/content/sessions/1170.md b/content/sessions/1170.md deleted file mode 100755 index 027ac6858..000000000 --- a/content/sessions/1170.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache Doris (incubating) -- A simple and single tightly coupled olap system" -date: "2019-09-10T17:00:00" -track: "bigdata" -presenters: "De Li" ---- - -Apache Doris (incubating) is an MPP-based interactive SQL data warehousing for reporting and analysis. It is open-sourced by Baidu. Doris mainly integrates the technology of Google Mesa and Apache Impala. Unlike other popular SQL-on-Hadoop systems, Doris is designed to be a simple and single tightly coupled system, not depending on other systems. Doris not only provides high concurrent low latency point query performance, but also provides high throughput queries of ad-hoc analysis. Doris not only provides batch data loading, but also provides near real-time mini-batch data loading. Doris also provides high availability, reliability, fault tolerance, and scalability. The simplicity (of developing, deploying and using) and meeting many data serving requirements in single system are the main features of Doris. \ No newline at end of file diff --git a/content/sessions/1171.md b/content/sessions/1171.md deleted file mode 100755 index 072c628c3..000000000 --- a/content/sessions/1171.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Inside Apache Druid: Built for High-Performance Real-Time Analytics" -date: "2019-09-10T18:00:00" -track: "bigdata" -presenters: "Surekha Saharan" ---- - -Interactive applications are replacing traditional reporting interfaces as the preferred means for organizations to derive value from their datasets. An interactive user experience requires latency on the order of milliseconds. Cluster computing frameworks such as Apache Hadoop or Apache Spark are tremendously beneficial in processing and deriving insights from data. However, high query latency makes these frameworks sub-optimal for interactive applications. Alternatively, the use of relational databases and key/value stores as dedicated query layers can reduce latency, but these approaches suffer many drawbacks for analytic use cases. Apache Druid (incubating) is well suited to power analytic applications working with real-time data and requiring low latency. This talk will cover the current state of analytics world, drawbacks of current popular databases, and what Druid nbrings to the table. The talk will cover technical details behind Druid’s design, including how its query processing layer works and how each component contributes to achieving top performance for analytical queries. We will discuss Druid modern architecture that implements a memory-mappable storage format, indexes, compression, late tuple materialization, and a query engine that can operate directly on compressed data. At the time of ApacheCon, Druid will very likely have been granted status as a top-level ASF project. \ No newline at end of file diff --git a/content/sessions/1172.md b/content/sessions/1172.md deleted file mode 100755 index ace4eb60f..000000000 --- a/content/sessions/1172.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache Training - Contributing more than just code" -date: "2019-09-10T17:00:00" -track: "community" -presenters: "Justin Mclean" ---- - -Does it seem strange to you that we collectively collaborate on code but training material is produced individually in private? Why would each company or person produce their own material when it can be sourced from a central location, under a business friendly license, and built on and modified? Or perhaps you just see better ways of producing content, then come along and listen to what the Apache Training project is doing. You'll find out how make nice presentations with simple markup that can be put under version control and exported to many formats. \ No newline at end of file diff --git a/content/sessions/1173.md b/content/sessions/1173.md deleted file mode 100755 index 05cf2dda8..000000000 --- a/content/sessions/1173.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "YuniKorn: A Universal Resource Scheduler for both YARN and Kubernetes" -date: "2019-09-10T11:15:00" -track: "bigdata" -presenters: "Wangda Tan" ---- - -'We will talk about our open source work - YuniKorn scheduler project (Y for YARN, K for K8s, uni- for Unified) brings long-wanted features such as hierarchical queues, fairness between users/jobs/queues, preemption to Kubernetes; and it brings service scheduling enhancements to YARN. Any improvements to this scheduler can benefit both Kubernetes and YARN community. Abstract: YARN schedulers are optimized for high-throughput, multi-tenant batch workloads. It can scale up to 50k nodes per cluster, and schedule 20k containers per second; On the other side, Kubernetes schedulers are optimized for long-running services, but many features like hierarchical queues, fairness resource sharing, and preemption etc, are either missing or not mature enough at this point of time. However, underneath they are responsible for one same job: the decision maker for resource allocations. We see the need to run services on YARN as well as run jobs on Kubernetes. This motivates us to create a universal scheduler which can work for both YARN and Kubernetes, and configure in the same way. This YuniKorn scheduler (Y for YARN, K for K8s, uni- for Unified) brings long-wanted features such as hierarchical queues, fairness between users/jobs/queues, preemption to Kubernetes; and it brings service scheduling enhancements to YARN. Most importantly, it provides the opportunity to let YARN and Kubernetes share the same user experience on scheduling big data workloads. And any improvements to this scheduler can benefit both Kubernetes and YARN community. In this talk, we’re going to talk about our efforts of design and implement the YuniKorn scheduler. We have integrated it with both YARN and Kubernetes. We will show demos and best practices.n' \ No newline at end of file diff --git a/content/sessions/1174.md b/content/sessions/1174.md deleted file mode 100755 index c403aa6bf..000000000 --- a/content/sessions/1174.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "First Step to Hybrid Cloud Computation: Elastic YARN and Kubernetes" -date: "2019-09-10T14:30:00" -track: "bigdata" -presenters: "Zhankun Tang, Sunil Govind" ---- - -'In this presentation, we will introduce our work: Elastic YARN, which is an elastic, cloud-native solution and it allows co-location YARN and Kubernetes for both on-premises and on-cloud. Audiences can learn how to build a hybrid, low-cost and easy-to-use computation platform while ensuring the SLA by leveraging our work. Abstract Enterprise continues to evolve towards data-driven, more types of workloads will come onboard, the challenges of managing mixed workloads emerge. It's essential to think about what is the most flexible architecture to manage batch jobs and long-running services both on-premises and on-cloud. Apache Hadoop YARN and Kubernetes are the de-facto enterprise level resource management system for big data workloads and long-running services. More specifically, big data applications like Spark/Hive-on-Tez/Map-Reduce are all prod-ready on YARN; on the other hand, Kubernetes is good to deploy services. Hadoop is running everywhere inside on-premises clusters, and Kubernetes is running everywhere inside clusters on the cloud. If a user wants to lift-and-shift workload from on-premises to on-cloud, it is still an extremely challenging task. In this presentation, we will introduce our work: Elastic YARN, which is an elastic, cloud-native solution and it allows co-location YARN and Kubernetes for both on-premises and on-cloud. Audiences can learn how to build a hybrid, low-cost and easy-to-use computation platform while ensuring the SLA by leveraging our work.n' \ No newline at end of file diff --git a/content/sessions/1175.md b/content/sessions/1175.md deleted file mode 100755 index 6626dea93..000000000 --- a/content/sessions/1175.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Wire Encryption In HDFS: Protect Your Data From Others, Not Yourself" -date: "2019-09-10T15:30:00" -track: "bigdata" -presenters: "Chen Liang" ---- - -Wire data encryption is a key component of the Hadoop Distributed File System (HDFS). HDFS can enforce different levels of data protection, allowing users to specify one based on their own needs. However, such enforcement comes in as an all-or-nothing feature. Namely, wire encryption is enforced either for all accesses or none. Since encryption bears a considerable performance cost, the all-or-nothing condition forces users to choose between 'faster but unencrypted' or 'encrypted but slower' for all clients. In our use case at LinkedIn, we would like to selectively expose fast unencrypted access to fully managed internal clients, which can be trusted, while only expose encrypted access to clients outside of the trusted circle with higher security risks. That way we minimize performance overhead for trusted internal clients while still securing data from potential outside threats. We re-evaluate the RPC encryption mechanism in HDFS. Our design extends HDFS NameNode to run on multiple ports. Depending on the configuration, connecting to different NameNode ports would end up with different levels of encryption protection. This protection then gets enforced for both NameNode RPC and the subsequent data transfers to/from DataNode. System administrators then need to set up a simple firewall rule to allow access to the unencrypted port only for internal clients and expose the encrypted port to the outside clients. This approach comes with minimum operational and performance overhead. The feature has been introduced to Apache Hadoop under HDFS-13541. \ No newline at end of file diff --git a/content/sessions/1176.md b/content/sessions/1176.md deleted file mode 100755 index 23b39d947..000000000 --- a/content/sessions/1176.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Ozone: Evolving HDFS Scalability to new heights & built-in GDPR Compliance" -date: "2019-09-10T17:00:00" -track: "bigdata" -presenters: "Dinesh Chitlangia" ---- - -Apache Hadoop Ozone is a robust, distributed key-value object store for Hadoop with layered architecture and strong consistency. It separates the namespace management from block and node management layer, which allows users to independently scale on both axes. Ozone is interoperable with Hadoop ecosystem as it provides OzoneFS (Hadoop compatible file system API), data locality and plug-n-play deployment with HDFS as it can be installed in an existing Hadoop cluster and can share storage disks with HDFS. Ozone solves the scalability challenges with HDFS by being size agnostic. Consequently, it allows users to store trillions of files in Ozone and access them as if they are on HDFS. Ozone plugs into existing Hadoop deployments seamlessly, and programs like Yarn, MapReduce, Spark, Hive and work without any modifications. In the era of increasing need for data privacy and regulations, Ozone also aims to provide built-in support for GDPR compliance with strong focus on Right to be Forgotten i.e., Data Erasure. At the end of this presentation the audience will be able to understand: 1. Overview of current challenges with HDFS scalability 2. How Ozone’s Architecture solves these challenges 3. Overview of GDPR 4. Built-in support for GDPR in Ozone \ No newline at end of file diff --git a/content/sessions/1177.md b/content/sessions/1177.md deleted file mode 100755 index e79bb7102..000000000 --- a/content/sessions/1177.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Hadoop Storage in the Cloud Native Era" -date: "2019-09-10T18:00:00" -track: "bigdata" -presenters: "Márton Elek" ---- - -Hadoop was born much earlier than the Cloud Native era. But the question is still the same: what can it offer in the time of Kubernetes, containerization and hybrid clouds? Apache Hadoop Ozone is a new subproject of Hadoop. It is based a generic low-level binary layer, the Hadoop Distributed Data Storage (HDDS) and a S3 compatible Object Store implementation on top of it. The separation of the data storage layer has multiple benefits: it can help to solve the “small files problem” and increase the scalability of the Hadoop storage clusters but it also helps us to provide access to the same data via multiple interfaces. The same data can be accessed via a Hadoop compatible file system API, with an S3 compatible file system API or as a raw, mountable block device. Storage is still a hot topic with Kubernetes and in Cloud Native environments. Container Storage Interface specification is a vendor neutral standard to provide storage plugin for multiple container orchestration system. With implementing the vendor-independent Container Storage Interface, Ozone can provide block level access on top of the Hadoop Distributed Data Storage layer and it can be first class citizen of the containerized word. In this talk we will demonstrate how the Apache Hadoop Ozone storage could be used from containers, how can the same storage space can be used as Object Store, Hadoop compatible file system or native block device. We will explain the basic storage type of Kubernetes clusters and show how Apache Hadoop Ozone can help to solve the storage problem in an industry standard way. \ No newline at end of file diff --git a/content/sessions/1178.md b/content/sessions/1178.md deleted file mode 100755 index 8b06f1034..000000000 --- a/content/sessions/1178.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Building BigData Query Optimization with Apache Calcite – Best Practices from Alibaba MaxCompute" -date: "2019-09-11T11:00:00" -track: "bigdata" -presenters: "Haisheng Yuan" ---- - -MaxCompute is a large scale, distributed big data platform of Alibaba, which provides Exabyte storage capacity and massive computing power through tens of thousands of commodity machines. The system supports SQL-like declarative language for advanced query and analysis on web-scale data set. Millions of jobs with hundreds of petabytes of data are processed every day, powering mission critical business within Alibaba, including e-Commerce, mobile payment, logistics, etc. 

Query optimizer plays a key part in determining the optimal execution plan. We first provide a general introduction about the overall architecture of MaxCompute, then we introduce how MaxCompute leverages Apache Calcite to build an efficient and robust query optimizer. We will also discuss the physical operators that MaxCompute creates in order to adapt to Apache Calcite, and the improvements that have been done to Calcite, such as IN list optimization, outer join null skew optimization, and logical relational node preprocessing. We further introduce our data partitioning techniques, including hash and range partitioning, to support advanced query parallelism. We will also talk about HBO (Historical Based Optimization) for regular ETL tasks. \ No newline at end of file diff --git a/content/sessions/1179.md b/content/sessions/1179.md deleted file mode 100755 index ac81ef299..000000000 --- a/content/sessions/1179.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Lessons Learned from Leveraging Real-Time Power Consumption Data with Apache Kudu" -date: "2019-09-11T12:00:00" -track: "bigdata" -presenters: "Masahiro Ito" ---- - -IoT is widely spreading in various industries, and various sensor devices generate large amounts of data in real-time. Such big data is used for visualization, machine learning, and other data analysis. Kudu is one of the data stores for such use cases which require fast inserts and efficient scans. Apache HBase is also often used to handle time series data such as sensor data. We also evaluated HBase for smart meter data a few years ago. HBase is suitable for storing sensor data because it has excellent insert performance, but it is not suitable for big data analysis because scan performance is not so good. On the other hand, Kudu is suitable for both fast insert and big data scan. We evaluated the performance of Kudu for leveraging the power consumption data collected in every second from sensor devices. With this performance test, we confirmed that 4-node Kudu cluster can storing and referring data in real-time generated by 1 million sensor devices. In order to handle a large amount of data in real-time, it is important to increase the memory allocation, to reduce the tablet size, and to define the order of the primary key to reduce the load of compaction. This presentation will introduce the differences between Kudu and other data stores (HBase, HDFS), the use cases suitable for Kudu, and the tuning know-how obtained through performance test. \ No newline at end of file diff --git a/content/sessions/1180.md b/content/sessions/1180.md deleted file mode 100755 index 3d33db704..000000000 --- a/content/sessions/1180.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Using Relational Cache to Boost Apache Spark SQL" -date: "2019-09-11T14:15:00" -track: "bigdata" -presenters: "Daoyuan Wang" ---- - -Apache Spark has become a popular engine for data analysis in industry, providing a handful SQL interface and processing data from various data sources. While many of our customers are using Apache Spark as an interactive query platform to meet their business requirement, one of the most common complaints we heard is that spark is not quite 'interactive' while data goes big. In traditional database systems, we can utilize materialized views to accelerate query processing, through pre-computation and query plan rewriting. We adopt a similar method to Spark, allow user to create flexible caches from a query or cube definition, and rewrite user queries at runtime to utilize pre-computed results. The cached data can persist to external data sources that Spark supports, or stay in memory, and updates automatically when new data ingested. nIn this talk, we'd like to take a deep dive into our design, and show real-world performance gain from our customers. This feature will be contributed to community. \ No newline at end of file diff --git a/content/sessions/1181.md b/content/sessions/1181.md deleted file mode 100755 index 90cf27881..000000000 --- a/content/sessions/1181.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "DataSketches - The Required Toolkit for the Analysis of Big Data" -date: "2019-09-11T15:15:00" -track: "bigdata" -presenters: "Lee Rhodes" ---- - -In​ the analysis of b​ig data there are often problem queries that don’t scale because they require huge compute resources to generate exact results, or don’t parallelize well. Examples include c​ount-distinct, ​quantiles, most frequent items, joins, matrix computations, and graph analysis. Algorithms that can produce accuracy guaranteed approximate answers for these problem queries are a required toolkit for modern analysis systems that need to process massive amounts of data​ quickly. For interactive queries there may not be other viable alternatives, and in the case of real­-time streams, these specialized algorithms, called stochastic, s​treaming, sublinear algorithms,​ or 's​ketches',​ are the only known solution. This technology has helped Yahoo successfully reduce data processing times from days to hours or minutes on a number of its internal platforms and has enabled subsecond queries on real-time platforms that would have been infeasible without sketches. This article provides a short introduction to sketching and to DataSketches, an open source library of a core set of these algorithms designed for large production analysis systems. \ No newline at end of file diff --git a/content/sessions/1182.md b/content/sessions/1182.md deleted file mode 100755 index 3626e2478..000000000 --- a/content/sessions/1182.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Samza 1.0: How we scaled stream processing at LinkedIn" -date: "2019-09-11T16:45:00" -track: "bigdata" -presenters: "Jagadish Venkatraman" ---- - -'At LinkedIn, we generate a staggering 2 trillion events to Kafka each day. nThe ingested events are processed by our streaming platform powered by Apache Samza. nWe currently have over 3000 applications in production that leverage Samza at LinkedIn. nThe use-cases include the LinkedIn feed, anomaly detection, combating fraud, profile notifications, realtime analytics and many more. In this talk, we will describe our journey in building and scaling a distributed stream nprocessing system over these years. Drawing from our experience running 24x7 applicationsnat LinkedIn, we motivate the key-challenges common across them, how we naddressed them and the lessons learnt along the way. Specific challenges include - massive scale, accuracy of results, multi-language support, ndeveloper productivity, high-performance data-access and fast recovery. Lastly, we will nalso share strategic areas for our future work in stream processing.' \ No newline at end of file diff --git a/content/sessions/1183.md b/content/sessions/1183.md deleted file mode 100755 index 19d52c8e6..000000000 --- a/content/sessions/1183.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Web-based Interactive Big Data Visualization" -date: "2019-09-11T17:45:00" -track: "bigdata" -presenters: "Wenli Zhang" ---- - -Data visualization is a challenging job to do in the Web-based environment due to the limited rendering efficiency of Web browsers. Rendering big data and providing user interaction in real time is even more challenging. nIn this speech, we are going to introduce how we made this possible with ECharts, a popular Web-based data visualization tool. Generally speaking, the solution lies in the following four aspects:n1. Simplifying shapes and rendering faster but with a similar result.n2. Optimizing code to improve rendering efficiency.n3. Applying stream rendering to give viewers a partial response as soon as possible.n4. Split data into several files and render in a pipeline to speed up rendering. \ No newline at end of file diff --git a/content/sessions/1184.md b/content/sessions/1184.md deleted file mode 100755 index a7b20e172..000000000 --- a/content/sessions/1184.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Game Changing Operational Visibility tools in Apache Cassandra 4.0" -date: "2019-09-11T12:00:00" -track: "cassandra" -presenters: "Chris Lohfink" ---- - -In this talk, we will review the new and upcomming tools for Apache Cassandra operations. This talk will go over some of the new and soon to be tools for operational visibility. Specific areas covered will include:n- Different operational tasks will be exposed with an HTTP interface with the sidecarn- Virtual tables provides visibility to the state of a node with cqlshn- New commands added to nodetool and JMX See how these improvements and additions will give operation teams faster diagnostics to assist in discovering problematic clients and data models. \ No newline at end of file diff --git a/content/sessions/1185.md b/content/sessions/1185.md deleted file mode 100755 index d37a277f9..000000000 --- a/content/sessions/1185.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Combining schema-on-read and schema-provisioning in Apache Drill" -date: "2019-09-11T16:45:00" -track: "drill" -presenters: "Aman Sinha" ---- - -The data generated from IoT devices, machine logs and similar sources is often semi-structured or unstructured. This poses a challenge for traditional schema-on-write systems that require a fixed schema up-front for querying. Modern analytic applications often have ad hoc usage patterns and require tremendous flexibility over how this data can be consumed. Further, they demand that the data be made available for querying soon after it has landed in their data platform – which may be a distributed file system or NoSQL database or something similar. In this talk I will first describe how Apache Drill’s innovative schema-on-read capability built into a distributed SQL query engine meets the demands of such applications. The raw data in various formats can be queried directly from either the distributed file system or other data sources such as NoSQL DBs without having to define a schema up-front. The Drill query planner (in conjunction with Apache Calcite) supports the ‘ANY’ type for columns which allows type validations. During execution, readers produce ‘RecordBatches’ where all rows in one RecordBatch have the same schema but across batches the schema may change. This provides the core foundation for the schema-on-read capability. Downstream SQL operators in the query pipeline perform run-time Java code generation by generating schema specific code. This code is compiled and executed on the JVM and based on this the operator produces output batches. In cases where Drill is not able to infer the schema correctly or when there are ambiguities, Drill has introduced schema provisioning to complement the schema-on-read. I will describe the functionality to define columns types and nullability, specify columns format and default values and control which columns are projected and their projection order. The schema can be specified declaratively as part of the query or in a separate file. Our expectation is that these complementary strategies will meet the demands of the modern analytic applications that require schema flexibility in addition to performance and scalability. \ No newline at end of file diff --git a/content/sessions/1186.md b/content/sessions/1186.md deleted file mode 100755 index 2d5a82113..000000000 --- a/content/sessions/1186.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "OFBiz in the fashion Industry, an omnichannel approach" -date: "2019-09-12T10:00:00" -track: "ofbiz" -presenters: "Nicola Mazzoni" ---- - -The fashion industry works to meet the demand for apparel and dictates thentrends for what should be worn.nThis presentation will describe why Apache OFBiz is a great platform fornbuilding an ERP for the fashion industry and how it can be enhanced tonsupport its specific requirements. The various implementation steps of anconcrete project will be reviewed:n- Development of a custom eCommerce/B2B Responsive and Seo compliantnapplication using the Color-Size matrix feature.n- Integration of third-party applications such as retail, legacy Erp,nsocial network, WMS, Business Intelligence, mailing system.n- 'Soft migration'. Progressive replacement of the external proprietarynsystem with OfBiz. The result is an OfBiz-centric, holonic Erp system. \ No newline at end of file diff --git a/content/sessions/1187.md b/content/sessions/1187.md deleted file mode 100755 index 1377adb9d..000000000 --- a/content/sessions/1187.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Kafka, Cassandra and Kubernetes at Scale - Real-time Anomaly detection on 19 billion events a day" -date: "2019-09-12T13:00:00" -track: "cassandra" -presenters: "Paul Brebner" ---- - -Apache Kafka, Apache Cassandra and Kubernetes are open source big data technologies enabling applications and business operations to scale massively and rapidly. While Kafka and Cassandra underpins the data layer of the stack providing capability to stream, disseminate, store and retrieve data at very low latency, Kubernetes is a container orchestration technology that helps in automated application deployment and scaling of application clusters. In this presentation, we will reveal how we architected a massive scale deployment of a streaming data pipeline with Kafka and Cassandra to cater to an example Anomaly detection application running on a Kubernetes cluster and generating and processing massive amount of events. Anomaly detection is a method used to detect unusual events in an event stream. It is widely used in a range of applications such as financial fraud detection, security, threat detection, website user analytics, sensors, IoT, system health monitoring, etc. When such applications operate at massive scale generating millions or billions of events, they impose significant computational, performance and scalability challenges to anomaly detection algorithms and data layer technologies. We will demonstrate the scalability, performance and cost effectiveness of Apache Kafka, Cassandra and Kubernetes, with results from our experiments allowing the Anomaly detection application to scale to 19 Billion anomaly checks per day. \ No newline at end of file diff --git a/content/sessions/1188.md b/content/sessions/1188.md deleted file mode 100755 index c38132790..000000000 --- a/content/sessions/1188.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Reduce your Storage Costs with Transient Replication and Cheap Quorums" -date: "2019-09-12T14:00:00" -track: "cassandra" -presenters: "Alex Petrov" ---- - -nIn eventually consistent systems, when a node failures or network partition occurs, we’re presented with a trade-off: to execute a request and sacrifice consistency or reject execution and sacrifice availability. In such system, quorums, overlapping node subsets guaranteeing at least one node to hold the most recent value, can be a good middle-ground. We can tolerate failures and loss of connectivity for some nodes while still serving latest results. Quorum-based replication schemes incur high storage costs: we have to store redundant values on several nodes to guarantee enough copies are going to be available in case of failure. It turns out that we do not have to store data on each replica. We can reduce storage and compute resources by storing the data only a subset of nodes, and only use the other nodes (Transient Replicas), for redundancy in failure scenarios. In this talk, we discuss Witness Replicas, a replication scheme used in Spanner and Megastore, and Apache Cassandra implementation of this concept, called Transient Replication and Cheap Quorums. (edited) \ No newline at end of file diff --git a/content/sessions/1189.md b/content/sessions/1189.md deleted file mode 100755 index 7bbacf426..000000000 --- a/content/sessions/1189.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Building Zhaopin's enterprise event bus based on Apache Pulsar" -date: "2019-09-12T17:30:00" -track: "bigdata" -presenters: "Jia Zhai, Penghui Li" ---- - -Zhaopin.com is a Chinese online recruitment services provider. As a bilingual job board, Zhaopin.com has one of the largest selections of real-time job vacancies in China. All recruitment advertisements are provided by prominent Chinese and foreign companies and enterprises throughout China. Zhaopin.com provides professional HR services to over 2.2 million clients and its average daily page views are over 68 million.nApache Pulsar was developed to address several shortcomings of existing messaging systems including a lot of enterprise features, message durability, and lower message latency. Zhaopin.com had built its enterprise event bus By using RabbitMQ for years. As the company grows, the amount of its data is getting larger and larger, and the usage scenarios of message systems are also become varied. The original RabbitMQ based architecture was hard to afford. Ultimately, it chose Apache Pulsar to replace RabbitMQ based architecture in 2018. Since Apache Pulsar was deployed at August 2018, The data amount in Pulsar is increased from 30 million messages per day to 6 billion+ messages. By learning from a lot of practical experiences, they summarised the best practices to using Pulsar as an enterprise event bus. Penghui and Jia detail how Apache Pulsar meet the requirements from Zhaopin.com. How Apache Pulsar is used in Zhaoping.com, and What is the best practices to using Pulsar as an enterprise event bus. Along the way, they will highlight the advantages of Apache Pulsar over the old system, and how high durability, high throughput and low latency of Apache Pulsar make it ideally suited for enterprise event bus. \ No newline at end of file diff --git a/content/sessions/1190.md b/content/sessions/1190.md deleted file mode 100755 index 3a97836c9..000000000 --- a/content/sessions/1190.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Streamlining Streaming System Management with Apache Heron" -date: "2019-09-12T10:00:00" -track: "bigdata" -presenters: "Josh Fischer" ---- - -Apache Heron (Incubating) is a real-time, distributed, fault-tolerant stream processing engine. In 2018 it was donated to the Apache Foundation and is used for processing data sets in a way that allows systems to react to data as it is created. The big data era is here and it has become more critical than ever that organizations are able to make decisions based on the data they own in an efficient and accurate way. However, making this data available when and where it is needed can be problematic for organizations, product managers, and development teams across the globe. This is where data streaming steps in. The streaming system approach comes with both advantages and disadvantages. Heron assists with managing the pain points of streaming by relieving the developer from needing to manually intervene in many cases. Heron provides a way to build better distributed systems by implementing a process based architecture built in modern programming languages. We will walk through how Heron can assist organizations to build these systems with specific focus on use cases, system architecture, and system observability. \ No newline at end of file diff --git a/content/sessions/1191.md b/content/sessions/1191.md deleted file mode 100755 index 5e6262ae8..000000000 --- a/content/sessions/1191.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "One SQL to Rule Them All – a Syntactically Idiomatic Approach to Management of Streams and Tables" -date: "2019-09-12T13:00:00" -track: "bigdata" -presenters: "Kenneth Knowles, Julian Hyde" ---- - -Apache Calcite is a data management framework that includes a SQL parser and query optimizer. It is used by many projects that implement SQL processing capabilities, including Apache Beam and Apache Flink. Over the last years, members of these three communities had many discussions about the semantics and syntax of 'Streaming SQL'. End of last year, we decided to formalize and summarize our views and ideas in paper that we submitted to the Industrial Track of the SIGMOD 2019 conference. The paper got accepted (http://sigmod2019.org/sigmod_industry_list). It presents a three-part proposal for integrating robust streaming into SQL, namely: n(1) time-varying relations as a foundation for classical tables as well as streaming data,n(2) event time semantics, n(3) a limited set of optional keyword extensions to control the materialization of time-varying query results. The paper shows how with these minimal additions it is possible to utilize the complete suite of standard SQL semantics to perform robust stream processing and motivates and illustrate these concepts using examples and describe lessons learned from implementations in Apache Calcite, Apache Flink, and Apache Beam. In this talk, we present our 'Syntactically Idiomatic Approach to Manage Streams and Tables'. \ No newline at end of file diff --git a/content/sessions/1192.md b/content/sessions/1192.md deleted file mode 100755 index 858e21541..000000000 --- a/content/sessions/1192.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Interactive querying of streams using Apache Pulsar" -date: "2019-09-12T14:00:00" -track: "bigdata" -presenters: "Boyang Jerry Peng" ---- - -As applications become more reliant on real-time data, streaming/messaging platforms have become more and more popular and crucial to any data pipeline. Currently, many streaming/messaging platforms are only used to access the most recent events from streams of data, however, there is tremendous value that can be unlocked if the full history of streams can be queried in an interactive fashion. Pulsar SQL is a query layer built on top of Apache Pulsar (a next-gen messaging platform), that enables users to dynamically query all streams, old and new, stored inside of Pulsar. Thus, users can unlock insights from querying both new and historical streams of data in a single system. Pulsar SQL leverages Presto and Apache Pulsar’s unique architecture to execute queries in a highly scalable fashion regardless of the number of partitions of topics that make up the streams. In this talk, we will examine the use cases and advantages of being able to interactively query events within an streaming messagingnplatform and how Pulsar enables users to do that in the most user-friendly and efficient manner. \ No newline at end of file diff --git a/content/sessions/1193.md b/content/sessions/1193.md deleted file mode 100755 index 11d46f5d0..000000000 --- a/content/sessions/1193.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Serverless Event Streaming with Pulsar Function: Use Cases and Best Practices" -date: "2019-09-12T15:30:00" -track: "bigdata" -presenters: "Xiaolong.Ran" ---- - -As data continues to evolve, there are more and more requirements for calculating data. From servers to VMS to containers to serverless, the computing framework is constantly evolving with the needs of users. How to provide users with a fast, easy-to-deploy computing framework has become a question for everyone to think about. In this talk, we will introduce how Pulsar provides a powerful computing power based on the serverless computing architecture. Apache Pulsar is a new generation of cloud messaging system and real-time processing platform. The message system is closely related to the real-time computing platform, but it is often loosely deployed and managed separately. Pulsar Functions, as the computing component of Pulsar, is a fusion and innovation of the message and computing platform in the serverless direction. Pulsar Functions provides support for multiple languages such as Go, Python, and Java; and runtimes such as Thread, Process, and Kubernetes. This provides great flexibility for user-written, run, and deploy features. Let users only use logic that cares about real calculations, without complicated configuration or management; it is easier to build a message-triggered flow platform. \ No newline at end of file diff --git a/content/sessions/1194.md b/content/sessions/1194.md deleted file mode 100755 index 9c2a37915..000000000 --- a/content/sessions/1194.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Event Sourcing with Spring Boot and Apache Kafka" -date: "2019-09-12T16:30:00" -track: "bigdata" -presenters: "Valentin Zickner" ---- - -In our global economy, businesses must be nimble and often have to adapt quickly. As a result, many businesses structure their teams in an Agile way to keep up with this demand. With Domain Driven Design, it’s possible to quickly modify applications to accommodate changing business needs and easily integrate with disparate third-party systems. But what if you were able to use historical data and analytics to enhance your applications capabilities? You may not always have all the necessary information in your CRUD database. With Event Sourcing, you are able to store new application events as well as existing events resulting in more robust applications. This makes it possible to change your application in ways you have never imagined before! This talk will show you how you can design your Event Sourcing based application with Spring Boot and Apache Kafka. \ No newline at end of file diff --git a/content/sessions/1195.md b/content/sessions/1195.md deleted file mode 100755 index 25914f5d5..000000000 --- a/content/sessions/1195.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Spoilt for Choice – Kafka Streams vs. KSQL for Stream Processing on top of Apache Kafka" -date: "2019-09-12T09:00:00" -track: "bigdata" -presenters: "Kai Waehner" ---- - -Apache Kafka is a de facto standard streaming data processing platform. It is widely deployed as event streaming platform. Part of Kafka is its stream processing API “Kafka Streams”. In addition, the Kafka ecosystem now offers KSQL, a declarative, SQL-like stream processing language that lets you define powerful stream-processing applications easily. What once took some moderately sophisticated Java code can now be done at the command line with a familiar and eminently approachable syntax.nThis session discusses and demos the pros and cons of Kafka Streams and KSQL to understand when to use which stream processing alternative for continuous stream processing natively on Apache Kafka infrastructures. The end of the session compares the trade-offs of Kafka Streams and KSQL to separate stream processing frameworks such as Apache Flink or Spark Streaming. \ No newline at end of file diff --git a/content/sessions/1196.md b/content/sessions/1196.md deleted file mode 100755 index 3feebf3a3..000000000 --- a/content/sessions/1196.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Hadoop Submarine Ecosystem: Bringing Machine Learning and Big Data world (YARN and Kubernetes) together " -date: "2019-09-12T14:00:00" -track: "machinelearning" -presenters: "Sunil Govind" ---- - -'Data Scientists focuses on developing ML models with applications such as TensorFlow / MXNet / Caffe / XGBoost and do not deep dive into the complexities of computing and storage to run ML/DL jobs. Today’s most of ETL processed data are stored in HDFS and cloud, and leveraging this data to design strong ML models is a big challenge for a data scientist. In Big Data ecosystem, most of the ETL and batch jobs are running on Spark and Hive which process/ingest data to the same data stores. Data scientists find this challenging in order to effectively utilize these big data workloads to develop an effective ML model. Hadoop Submarine (https://hadoop.apache.org/submarine/) helps to bring these two worlds together and provides seamless integration across. Ecosystem around Hadoop Submarine helps to design and run ML workloads from a notebook. Integrations with notebook such as Zeppelin and workflow scheduler like Azkaban, helps user to consume data from spark or hive and effectively run ML jobs with ease on any compute cluster. In this deep dive session, we will demo the simplicity of Submarine by running distributed deep learning/machine learning applications on YARN and Kubernetes as simple as running it locally. We will also showcase the community effort in developing Submarine’s ecosystem which eases integrations with Zeppelin and Azkaban. Submarine project could easily launch in the same cluster to run DL/ML jobs without any additional upgrades or complexities of having different machines.n' \ No newline at end of file diff --git a/content/sessions/1197.md b/content/sessions/1197.md deleted file mode 100755 index 9b5795be4..000000000 --- a/content/sessions/1197.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "From (Single) ISP-CDN - To Multi-ISP CDN" -date: "2019-09-09T09:00:00" -track: "contentdelivery" -presenters: "Nir Sopher" ---- - -'Traffic-control was built as an ISP CDN, serving content within a single ISP. Using traffic-control for creating a multi-ISP CDN, puts up challenges in various aspects: starting from the deployment of the different TC components outside of the ISPs, towards the isolation between different ISPs, as well as self-service, dealing simultaneously with different structures of ISP networks, the changing popularity of content in different ISP etc. In this session we would like to described few of the above challenges, as well as the derived traffic-control components requirements, flexibility and modularity.' \ No newline at end of file diff --git a/content/sessions/1198.md b/content/sessions/1198.md deleted file mode 100755 index ddb88a689..000000000 --- a/content/sessions/1198.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Let's Encrypt Apache Traffic Control" -date: "2019-09-09T14:00:00" -track: "contentdelivery" -presenters: "Matt Jackson" ---- - -Let's Encrypt is a free, automated certificate authority (CA) which is open for use to anyone who owns a domain name. Using Traffic Router as the DNS, we have integrated with Let's Encrypt when utilizing HTTPS on the delivery service domains. Many open source projects provide libraries to ease implementation of Let's Encrypt, and for this project we used go-acme/lego (MIT licensed) to integrate with the Traffic Ops Golang API. Integration with Let's Encrypt provides many advantages including saving money by using a completely free CA, saving time with automation including certificate renewal, and improving reliability by eliminating manual intervention for certificate generation and deployment. \ No newline at end of file diff --git a/content/sessions/1199.md b/content/sessions/1199.md deleted file mode 100755 index e414b8c8e..000000000 --- a/content/sessions/1199.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Processing CDN Logging Events at Scale" -date: "2019-09-10T17:00:00" -track: "contentdelivery" -presenters: "Geoff Genz" ---- - -'Content Delivery Networks generate staggering amounts of event data at a rate that is exponentially increasing in line with the explosion of IP video. Managing this deluge with the usual suspects of Big Data platforms presents a formidable challenge, particularly in light of the often-competing goals of detailed troubleshooting investigation versus big picture analytics and machine learning. This presentation will explore capturing and processing events from the largest production deployment of Apache Traffic Server and Apache Traffic Control. We will examine some of the drawbacks of standard approaches, including Splunk, the Elastic stack, Hadoop, and time series and OLAP datastores. This will be followed by a deep dive into Comcast’s rapidly evolving MAPLE platform. MAPLE leverages the open source Clickhouse “warp speed” database, along with Apache staples Kafka and Zookeeper, to collect and transform more than two million CDN events per second. This firehose of data is available within seconds at both an individual event and aggregate level for alerting, troubleshooting, analytics, and machine learning.n' \ No newline at end of file diff --git a/content/sessions/1200.md b/content/sessions/1200.md deleted file mode 100755 index 5e09145f9..000000000 --- a/content/sessions/1200.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "A Mile in My Shoes: How Empathy & Compassion Can Transform Your Work & Your Life" -date: "2019-09-09T16:30:00" -track: "community" -presenters: "Ruth Holloway" ---- - -In many places, we're told to 'leave our emotions out of it,' in our work, and in our other technical endeavors, including in open source communities. But that is an irrational, dangerously naïve point of view, one that often leads us to not concern ourselves with the feelings of others, thereby allowing biases (intended or not) to come into play. This harms our communities, by excluding valuable contributors, and their skills and ideas. Empathy and compassion are powerful tools. Taking the time to understand the though processes of another person, and the variety of struggles that we all have, lets you serve them, collaborate with them, and lead them more effectively than you've ever done before. In this talk, you'll learn the science behind these powerful emotions, and how using them can help you build stronger, more inclusive communities. Empathy and compassion have the power to enrich your life, and that of the people you interact with ever day. Come learn how you can use these vital skills to work more effectively and transform your world into a more caring, open, and peaceful place to be. \ No newline at end of file diff --git a/content/sessions/1202.md b/content/sessions/1202.md deleted file mode 100755 index 9fab0c864..000000000 --- a/content/sessions/1202.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Enterprise Hybrid Mobile App Development" -date: "2019-09-12T10:00:00" -track: "mobile" -presenters: "Mary Grygleski" ---- - -This session shows how to develop and run native mobile apps without needing to know much about the native platforms, such as Android and iOS, by leveraging a hybrid development framework such as Apache Cordova, which is capable of transforming an ordinary web app into a mobile app. The presentation discusses a real-life enterprise-level Android and iOS mobile app developed under pressure for a wireless service provider. It covers some of the technical details of the app as well as how to perform device-level debugging of the Android app, using adb. You’ll get practical tips on what to watch out for and compare the hybrid approach to React and Kotlin/Native \ No newline at end of file diff --git a/content/sessions/1203.md b/content/sessions/1203.md deleted file mode 100755 index f34ecd167..000000000 --- a/content/sessions/1203.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Advanced Apache Cordova Development" -date: "2019-09-12T13:00:00" -track: "mobile" -presenters: "Hazem Saleh" ---- - -Apache Cordova is a platform for building mobile apps using common Web technologies (HTML, CSS and JavaScript). Apache Cordova offers a set of APIs that allow the mobile app developers to utilize mobile native functions such as (Audio, Camera, Contacts …etc) using JavaScript. This session discusses and demonstrates the advanced aspects of Apache Cordova. Advanced aspects include extending Apache Cordova framework by creating custom plugins in order to access more device hardware features. Advanced aspects also include empowering Apache Cordova apps with Push notifications and Cloud services. Finally, this session demonstrates the power of automated JavaScript unit testing for producing quality Apache Cordova apps. \ No newline at end of file diff --git a/content/sessions/1204.md b/content/sessions/1204.md deleted file mode 100755 index ee98ec7ab..000000000 --- a/content/sessions/1204.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache Fineract CN Mobile 2.0 - Supporting Branchless banking and promoting financial inclusion in the fragile states" -date: "2019-09-12T14:00:00" -track: "mobile" -presenters: "Rajan Maurya" ---- - -In today's world, we don't have an offline first open source mobile banking solution. Nowadays It is too important to have an offline first application because in many rural areas, it is exceptionally hard to find an internet connection, sometimes taking days to get connected if a natural disaster happens Still, there are some areas where internet connection is still not even present and despite the many global efforts to bring connectivity forward, there are regions where there is still no timeline for connectivity. On the bright side, people can get access to an internet connection but they have to travel for that. Apache fineract-cn-mobile 2.0 is a native Android application built on the top of Apache Fineract-CN. Apache Fineract CN is an Application Framework for Digital Financial Services. It is a system to support nationwide financial transactions and to support the creation of an inclusive, interconnected digital economy for every nation in the world. Mobile devices are now in the palm of almost every hand worldwide, some high-speed internet connection and some with no internet or very slow internet connection in many places. I present Apache fineract-cn-mobile 2.0 that is the open source offline-first mobile banking application build on top of Apache Fineract-CN and maintained with Android Jetpack. Now we have eliminated as a barrier, no network or slow internet connection. The codebase is very simple and easy to fix and debug. Typically as a project grows, so does its complexity, but with Android Jetpack and MVVP architecture, complexity remains nearly a constant as when it was written in the first implementation. Fineract-cn-mobile 2.0 solves the topmost problem in financial institutions application development.n1. Maintainability and scalability with MVVM architecture and Android Jetpack (This give us the ability to adopt a new architecture that will come in future better than MVVM ).n2. Sustainability with ease of data synchronization.n3. Battery Optimizatio 4. Offline first, No need for internet connection. WorkManager jobs take care of data sync to the server and from the server.n5. Easy learnable UI & UX.n6. Easy SQLite local database management with Roomn7. Logic abstraction with MVVP (Model View ViewModel)n8. Updating UI with LiveData I will be talking about the above problems that we solved with latest Android tech. I will guide developers on how they can pick up the codebase and can develop high-standard features according to his/her need and can manage customers entirely from Android device and How fineract-cn-mobile helps alleviate poverty and makes banking paperless. \ No newline at end of file diff --git a/content/sessions/1205.md b/content/sessions/1205.md deleted file mode 100755 index cabb4772d..000000000 --- a/content/sessions/1205.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "The Power of Plugins - Developing plugins for Apache Cordova" -date: "2019-09-12T15:30:00" -track: "streaming" -presenters: "Jesse MacFadyen" ---- - -Plugins are the powerhouse inside Apache Cordova applications, this isnwhere all native functionality is exposed, and your mobile app trulynbecomes hybrid. We will explore the full scope of plugins in ApachenCordova, not just how they work, but how they are created, installed,ndiscovered and distributed. We will deep dive into concrete examples tondemonstrate the power of plugins. \ No newline at end of file diff --git a/content/sessions/1206.md b/content/sessions/1206.md deleted file mode 100755 index bd35f2b58..000000000 --- a/content/sessions/1206.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "OpenWhisk and Kubernetes based Mobile Backend as a Service" -date: "2019-09-12T16:30:00" -track: "mobile" -presenters: "Marek Sadowski" ---- - -So you've been hearing a lot of buzz about Serverless tech in conjunction with Containers and Mobile, but what exactly is the OpenWhisk in reference to Kubernetes? Come find out at this session. Serverless has become the new style of coding, and it might be perfect to offload your mobile apps, and container based systems without incurring unnecessary costs.nGet an intro into serverless/function-as-a-service/cloud native technologies in the Mobile-Backend-as-a-Service (MBaaS) context and learn why startups and enterprises are so excited about using it. We will be demoing Serverless and containers in Swift and iOS, and in Java for Android, and covering the following:n- Who are the players in the serverless ecosystemn- What are some use cases for serverless and container based solutions - with MBaaS as one of themn- Best practices for the serverless with the container system architecture for MBaaSn- If going serverless is really faster, better, cheaper for developers and organizationsn- Live coding examples using Swift and Java \ No newline at end of file diff --git a/content/sessions/1207.md b/content/sessions/1207.md deleted file mode 100755 index d36f60ae1..000000000 --- a/content/sessions/1207.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Tips and Tricks - Getting the most out of Apache Cordova" -date: "2019-09-12T17:30:00" -track: "streaming" -presenters: "Jesse MacFadyen" ---- - -Mobile development is hard, no matter what tools you use. With thenmultitude of devices out there how can you maximize the chance to delightnyour users? Jesse has been actively contributing to hybrid mobilendevelopment projects and writing hybrid mobile apps for 10 years and willnshare his bag of tricks. Learn best practices so you can optimize yournworkflow, and develop the best possible mobile applications. \ No newline at end of file diff --git a/content/sessions/1208.md b/content/sessions/1208.md deleted file mode 100755 index 5cdd6a38f..000000000 --- a/content/sessions/1208.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache Karaf on cloud, the kloud initiative" -date: "2019-09-12T14:00:00" -track: "karaf" -presenters: "JB Onofré" ---- - -Apache Karaf has been seen for a while as an applications server. If it's actually the case for the standard distribution, it's very easy to create a Karaf powered standard/immutable distribution embedding your applications. In combination with other Karaf subprojects (Decanter, Cave, ...), it provides a perfect applications runtime for the cloud.nThis talk will show how to start and focus on business code and then generate different distribution packages, up to a docker image ready to run on cloud. \ No newline at end of file diff --git a/content/sessions/1209.md b/content/sessions/1209.md deleted file mode 100755 index 74c8a98a8..000000000 --- a/content/sessions/1209.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Serverless: Multi-tenant Rule Engine Service Powered by Apache Karaf" -date: "2019-09-12T15:30:00" -track: "karaf" -presenters: "Dmitry Vasilyev, Saeid Mirzaei, George Ye" ---- - -The Netflix media pipeline processes thousands of new shows and movies every day so that you can watch them on any device anywhere. We use a forward chaining rule engine to coordinate all of this work in multiple workflows. Hosting these workflows in a reliable, scalable and cost effective manner is a huge challenge at our scale. In this talk, we will introduce the design of Netflix’s next generation rule engine framework. The goal is to boost modularity, increase developer productivity and decrease operational overhead. The new system is a platform as a service that lets workflow developers focus on workflow data model, execution conditions, and remote function invocations without worrying about how to deploy, scale, and monitor it. The system uses the OSGI framework to build separation among workflows and leverages Apache Karaf as the runtime container. Other interesting topics such as workflow bundle management and a novel rule domain specific language will be covered in this talk. Keywords:nRule engine, OSGI, Apache Karaf, Serverless \ No newline at end of file diff --git a/content/sessions/1210.md b/content/sessions/1210.md deleted file mode 100755 index 6f7caf7f8..000000000 --- a/content/sessions/1210.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Customer segmentation and personalization in websites/PWAs using Apache Unomi" -date: "2019-09-12T16:30:00" -track: "karaf" -presenters: "Serge Huber" ---- - -In this session, you will learn all that’s new with Apache Unomi, the open source Customer Data platform (that graduated this year) based on the Apache Karaf runtime, and all that’s happened since the last ApacheCon. You will discover how to easily integrate it with an existing website or SPA/PWA using its built-in web tracker, how to build customer segments and how to use the API to personalize the experience for your users. You’ll also learn how you can extend it to do almost anything, using either the built-in rules engine or your own plugins. You will also discover the new Docker compatibility and the upcoming GraphQL API. Finally, you’ll learn what’s next and how you can help the project. \ No newline at end of file diff --git a/content/sessions/1211.md b/content/sessions/1211.md deleted file mode 100755 index 5de11f03e..000000000 --- a/content/sessions/1211.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Testing Contributions at Scale" -date: "2019-09-10T18:00:00" -track: "general" -presenters: "Allen Wittenauer" ---- - -Time may be one of the most valuable resources in a project. Automating code reviews to allow for other tasks is a crucial goal for many communities. This talk will cover one way many Apache projects have significantly increased code quality and contribution feedback while simultaneously doing more with less. \ No newline at end of file diff --git a/content/sessions/1212.md b/content/sessions/1212.md deleted file mode 100755 index 0a7581db2..000000000 --- a/content/sessions/1212.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Peeking Behind the Curtain: The Anatomy of a Real Major Incident" -date: "2019-09-10T12:15:00" -track: "general" -presenters: "George Miranda" ---- - -Failures are inevitable. But when they occur, our goal should be to resolve them as quickly and efficiently as possible. PagerDuty has developed an open-source Incident Response framework based on the Incident Command System (ICS). That product-independent process has helped many organizations set up Incident Response processes that resolve technical issues as quickly and effectively as possible. There's a lot of documentation you can follow, but how do incidents actually play out in real-time when they happen? In this talk, you get to peek behind the curtain to see what happens inside the walls of the company that's known for waking you up to tell you there's a technical incident you need to deal with. We will walk through a cascading failure that happened when unexpected errors were found happening in one of our Kafka clusters. I share all the gritty details of this actual incident, as they occurred, and use that as a way to demonstrate the structure of how we apply ICS to resolving technical problems. Attendees of this talk will walk away with an understanding of how to effectively manage complex technical incidents across multiple teams, the additional roles that are necessary to support technical responders, how to structure a blameless post-mortem, and how to start developing a similar process in their own organizations. This talk delves into technical concepts due to the nature of the problem, but it is mostly focused on the mechanics of managing any technical incident. \ No newline at end of file diff --git a/content/sessions/1213.md b/content/sessions/1213.md deleted file mode 100755 index b560da80e..000000000 --- a/content/sessions/1213.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "New Features of Apache RocketMQ" -date: "2019-09-09T09:00:00" -track: "general" -presenters: "Von gosling, Heng Du" ---- - -In the past year, Apache RocketMQ has released a series of new features including transactional messages, ACLs, message track, multiple replicas storage, and multi-language clients. These features are all from contributors from many companies. nThis talk is not only about bringing technical analysis related to these new features, but also will share how to collaborate on the development of important features in the community, which is very important for Apache. \ No newline at end of file diff --git a/content/sessions/1214.md b/content/sessions/1214.md deleted file mode 100755 index 0322fd598..000000000 --- a/content/sessions/1214.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Java + Serverless - Finally a match made in Heaven?" -date: "2019-09-09T10:00:00" -track: "general" -presenters: "Pratik Patel" ---- - -Why isn’t Java a more popular language for serverless developers? How does Apache Openwhisk work with Java? Despite there being 10 million Java developers, Java is being used for only 6% of serverless workloads according to a recent survey. Historical problems with long cold-starts, heavy-weight frameworks and lack of tooling have made Java an unpopular choice for serverless development…. until now! In this session, I will take you on a journey to explain what has changed with Java to finally make it an amazing language for building serverless applications, using the open source Apache Openwhisk platform! In this session, we cover these topics including… Previous issues Java developers faced using serverless platformsnWhy JVMs struggled with serverless execution environments.nWhat makes Serverless Java now possible:nJVM advancesnLow-overhead, serverless ready frameworksnAOT (ahead of time) compilation and no-VM binariesnThis session will end with an earth-shaking demo: a full Java application that includes ORM, Dependency Injection, Annotations, etc, that starts in less than 50ms! \ No newline at end of file diff --git a/content/sessions/1302.md b/content/sessions/1302.md deleted file mode 100755 index b311ac2b4..000000000 --- a/content/sessions/1302.md +++ /dev/null @@ -1,15 +0,0 @@ ---- -title: "Mythbusting with Apache Cassandra" -date: "2019-09-11T15:15:00" -track: "cassandra" -presenters: "Jeffrey Carpenter" ---- - -As a mature operational database with a 10+ year history as an Apache project, Cassandra has a long history with many twists and turns. In this talk, we’ll look at common misgivings and misunderstandings that we encounter in conversations with developers and get to the bottom of the story: - -- Cassandra is a “columnar” (or “column-oriented”) database. -- Porting my relational data model to Cassandra will work well. -- Cassandra is a niche database that is only good for narrow set of use cases. -- My application can’t tolerate eventual consistency. -- Cassandra counters are unreliable. -- Never use materialized views or secondary indexes in production \ No newline at end of file diff --git a/content/sessions/1304.md b/content/sessions/1304.md deleted file mode 100755 index 8ac704953..000000000 --- a/content/sessions/1304.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Portable Spark Runner: Running Beam Pipelines Written in Python and Go with Spark" -date: "2019-09-11T11:00:00" -track: "beam" -presenters: "Kyle Weaver, Ismaël Mejía" ---- - -Apache Spark is the most popular open source analytics engine for large-scale data processing. Spark is not only a mature system, but thanks to its support of multiple resource managers like Hadoop, Mesos, and Kubernetes it has become a popular choice for both batch and streaming workloads in the industry. Apache Beam has included a Spark runner since its inception to allow users to execute Beam pipelines on Spark, but until recently the Spark runner could only execute pipelines written in Java. In this talk we will introduce the portability framework and how we adapted it into the existing Spark runner translation to make the Spark runner portable. We will show you how to execute Beam pipelines written in Python and Golang in Spark with Beam and invite you to use the new Spark Portable Runner. We will mention the use case of Tensorflow Extended, the end-to-end platform for data validation and transformation and ML model analysis. Finally we will discuss ongoing work and some future plans for the portable runner. \ No newline at end of file diff --git a/content/sessions/1305.md b/content/sessions/1305.md deleted file mode 100755 index d395dbe7d..000000000 --- a/content/sessions/1305.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Friendlier Communities Through Infrastructure: Apache Beam’s Journey" -date: "2019-09-11T12:00:00" -track: "beam" -presenters: "Jason Kuster, Alan Myrvold" ---- - -In this talk we will discuss the infrastructure Apache Beam has built to create a positive contributor experience. Over the last few years we have developed tooling, refactored code, instrumented Jenkins builds, collected metrics, created a metrics dashboard, and involved the community in order to ensure happy contributors. We will discuss the importance of engineering productivity and tooling in developing strong communities, where we have succeeded, and where we still have work to do. \ No newline at end of file diff --git a/content/sessions/1306.md b/content/sessions/1306.md deleted file mode 100755 index 536f1ac08..000000000 --- a/content/sessions/1306.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Using the Timer and State API to solve times series use case in Apache Beam" -date: "2019-09-11T14:15:00" -track: "beam" -presenters: "" ---- - -Time series processing requires ordered processing of data points, often across windows boundaries. Tricky stuff for distributed processing systems, in this chat we look at how to solve for these problems using a combination of Global Windows, State and the Timer API' with Apache Beam. \ No newline at end of file diff --git a/content/sessions/1307.md b/content/sessions/1307.md deleted file mode 100755 index 1c57e9300..000000000 --- a/content/sessions/1307.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "The journey of building a Beam runner based on Spark structured streaming framework" -date: "2019-09-11T15:15:00" -track: "beam" -presenters: "Etienne Chauchot" ---- - -Apache Beam provides a unified programming model to execute batch and streaming pipelines on all the popular big data engines. The translation layer from Beam to the chosen big data engine is called a runner. The current runner for Apache Spark is based on the RDD/DStream framework. However, there is an ongoing work to move it to Spark next generation framework a.k.a structured streaming. This talk will present why structured streaming is a good fit for Apache Beam, why it is worth the effort, and will give some feedback on how Apache Beam has solved the challenge, what the tough points and the sweet points were. \ No newline at end of file diff --git a/content/sessions/1308.md b/content/sessions/1308.md deleted file mode 100755 index 5926ae200..000000000 --- a/content/sessions/1308.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Developing new IO connectors in Apache Beam" -date: "2019-09-11T16:45:00" -track: "beam" -presenters: "Alexey Romanenko, Ismaël Mejía" ---- - -One of the key components in any data processing system is IO connectors. These fundamental blocks allow to read and write data, which is stored in different type of sources, in a unified and distributed way. In this sense, Apache Beam is not an exception - it provides a rich API to develop a new connector with your favorite SDK and easily integrate it with Beam runners. In this talk we are going to show you how to write your own IO connector (in Java). We will see what are the differences between bounded and unbounded sources, how to implement efficient sources and sinks and where we need to pay more attention during the development, what kind of API your connector should provide to users and how it can be tested. To achieve this we will rely on examples from existing Beam connectors. We will also give a brief overview of a rather recent feature/pattern in Beam IO - Composable IO connectors and the Splittable DoFn API. We will discuss the advantages of modular IO API design and some new IO design patterns allowed by this style. This talk will be interesting for people, who consider to write their own IO connectors or want to contribute to existing ones, as well as for Beam users, who wish to understand how existing Beam connectors work under the hood. \ No newline at end of file diff --git a/content/sessions/1309.md b/content/sessions/1309.md deleted file mode 100755 index c77e6d6cc..000000000 --- a/content/sessions/1309.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Simple, Portable data pipelines with Apache Beam SQL" -date: "2019-09-11T17:45:00" -track: "beam" -presenters: "Andrew Pilloud" ---- - -Apache Beam is a unified data processing framework, allowing you to write batch and streaming pipelines that run anywhere, including Apache Flink, Apache Spark, and Google Cloud Dataflow. With the SQL extension you can now write a pipeline in pure SQL. If you need more, you can write user defined functions in Java or even embed SQL into your existing Java pipeline. This talk will start with a demo pipeline written in pure SQL. We will review how streaming SQL came from collaboration between the Apache Beam, Apache Calcite and Apache Flink communities. Finally, we will deep-dive into the architecture of Beam’s implementation and the work we are doing to make Apache Beam SQL the default choice for writing new streaming pipelines. \ No newline at end of file diff --git a/content/sessions/1316.md b/content/sessions/1316.md deleted file mode 100755 index f85fe9696..000000000 --- a/content/sessions/1316.md +++ /dev/null @@ -1,9 +0,0 @@ ---- -title: "Running Apache Flink and Apache Beam on Kubernetes" -date: "2019-09-12T09:00:00" -track: "beam" -presenters: "Micah Wylde" ---- - -Access to real-time data is increasingly important for many organizations. At Lyft, we process millions of events per second in real-time to compute prices, balance marketplace dynamics, detect fraud, among many other use cases. To do so, we run dozens of Apache Flink and Apache Beam pipelines. Flink provides a powerful framework that makes it easy for non-experts to write correct, high-scale streaming jobs, while Beam extends that power to our large base of Python programmers. -Historically, we have run Flink clusters on bare, custom-managed EC2 instances. In order to achieve greater elasticity and reliability, we decided to rebuild our streaming platform on top of Kubernetes. In this session, I'll cover how we designed and built an open-source Kubernetes operator for Flink and Beam, some of the unique challenges of running a complex, stateful application on Kubernetes, and some of the lessons we learned along the way. \ No newline at end of file diff --git a/content/sessions/1317.md b/content/sessions/1317.md deleted file mode 100755 index 19b640e1f..000000000 --- a/content/sessions/1317.md +++ /dev/null @@ -1,10 +0,0 @@ ---- -title: "Driving dynamic Beam pipelines" -date: "2019-09-12T10:00:00" -track: "beam" -presenters: "Alex Van Boxel" ---- - -Using Apache Beam to get data in your data lake? In a agile company you don’t want to re-compile your ingestion pipeline every time a sprint finished. In this talk we go over all mechanisms and building blocks you need to make dynamic pipelines really work. - We’ll see why schemas are so important. How do we get these schemas in our pipelines and discuss methods to protect ourselves from data corruption and incompatible schema evolution. - The new features like schema aware PCollection get a thorough deep dive and finally we go over real world examples and position Apache Beam in the new PLT (Push Load Transform) world. \ No newline at end of file diff --git a/content/sessions/1318.md b/content/sessions/1318.md deleted file mode 100755 index 099ab08c4..000000000 --- a/content/sessions/1318.md +++ /dev/null @@ -1,9 +0,0 @@ ---- -title: "A Guided Walkthrough of the Apache Beam Go SDK" -date: "2019-09-12T13:00:00" -track: "beam" -presenters: "Robert Burke" ---- - -This session will walk through the internals of what the Go SDK does to execute a pipeline. Describing the structure of the SDK, where code is located, and the why of the structure, and how it does it with the particular features of the Go language, which notably has strict types but no generics. -Particular effort will be spent on what the Go SDK does for ease of use, and to reduce DoFn execution overhead. \ No newline at end of file diff --git a/content/sessions/1319.md b/content/sessions/1319.md deleted file mode 100755 index 47257e008..000000000 --- a/content/sessions/1319.md +++ /dev/null @@ -1,11 +0,0 @@ ---- -title: "Apache Beam + Kotlin = ❤️ " -date: "2019-09-12T14:00:00" -track: "beam" -presenters: "Harshit Dwivedi" ---- - -Kotlin is an Open Sourced, statically typed language for JVM and is mostly being favoured by Android Developers due to the many myriad features which enable more concise and cleaner code than Java without sacrificing performance or safety. - By using Kotlin as our language of choice to build an apache beam pipeline, our development team not only ran into fewer errors than what we would have if we used Java; but with the help of extensively useful high level APIs present in Kotlin, we were also able to cut down on the development time required to build, test and deploy a new feature. - This talk uses real life examples from my personal experiences to advocate on the benefits that you can get by using Kotlin to build solutions in Apache Beam. - So if you are someone who is already working with Apache Beam's Java SDK, or looking forward to try Apache BEAM, attending this talk will motivate you to go ahead and explore the benefits associated with using Beam with Kotlin. \ No newline at end of file diff --git a/content/sessions/1320.md b/content/sessions/1320.md deleted file mode 100755 index 2fc8ddddb..000000000 --- a/content/sessions/1320.md +++ /dev/null @@ -1,12 +0,0 @@ ---- -title: "Unifying Batch and Stream Data Processing with Apache Calcite and Apache Beam" -date: "2019-09-12T15:30:00" -track: "beam" -presenters: "Khai Tran" ---- - -At LinkedIn, we have a large and ever-expanding corpus of big data processing code written using batch processing frameworks like Pig, Hive, and Spark. These data processing pipelines produce derived data artifacts like metrics, dimensions, and features and are generally run at daily or hourly end-to-end latencies. For a subset of these artifacts, there often arises a need to produce them at a faster cadence, say minutely or even continuously. At that point, the common practice used to be that the developer would re-write the same derivation logic using a stream processing framework (Apache Samza in case of LinkedIn). This causes two problems: - 1. The time to go from a pipeline that generates data at hourly grain to a pipeline that generates the same data continuously is governed by developer time to understand the batch code and rewrite it in streaming language; typically weeks to a month to fully operationalize this pipeline. - 2. The rewritten pipeline may not be exactly the same as the batch pipeline because of subtle differences in the logic or UDFs across the two code bases. Over time these pipelines can continue to diverge especially if different teams own the batch and streaming pipelines. - In this talk, we present our solution to this batch-stream divide; a system that can auto-generate streaming code from batch logic and employ the Lambda architecture to transparently deliver merged results to data consumers. Technically, our Lambda architecture is very similar to the traditional Lambda architecture; however, our users only need to maintain a single code base in batch logic. This code base serves as a single source of truth to specify what users want to compute. If they need fresher results, they just need to turn on the streaming flag. - We will deep-dive into technical details for this system which is built using open source components: Apache Beam and Apache Calcite. We’ll also share our experience using this system to auto-migrate batch scripts in our unified metrics platform into near realtime pipelines running on Apache Samza and our plans in the future to share this with the community. \ No newline at end of file diff --git a/content/sessions/1321.md b/content/sessions/1321.md deleted file mode 100755 index 08abf22d7..000000000 --- a/content/sessions/1321.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Samza Portable Runner for Beam" -date: "2019-09-12T16:30:00" -track: "beam" -presenters: "Hai Lu" ---- - -In this talk we are going to cover how we have leveraged portability of Beam and make Stream Processing in Python possible on top of Apache Samza. We will first touch points on Apache Samza in general, and how it stands out as the stream processing engine at LinkedIn that scales to over a trillion messages processed per day, with strong state support and flexible deployment models. Next we introduce Samza Runner for Beam, particularly the portable runner. We will cover the efforts we have done to leverage Beam and make stream processing in Python available at LinkedIn, followed by a few use cases. The talk will conclude with our plan for future work. \ No newline at end of file diff --git a/content/sessions/1322.md b/content/sessions/1322.md deleted file mode 100755 index a0f17fc6b..000000000 --- a/content/sessions/1322.md +++ /dev/null @@ -1,16 +0,0 @@ ---- -title: "Creating a Stream Data Pipeline on Google Cloud Platform using Apache Beam" -date: "2019-09-12T17:30:00" -track: "beam" -presenters: "Shuichi Suzuki" ---- - -We built a scalable and flexible stream data pipeline for our microservices on Google Cloud Platform (GCP), using Cloud Pub/Sub, Google Cloud Storage, BigQuery, and Cloud Dataflow, using Apache Beam. The stream data pipeline is working on the production system for Mercari, one of the biggest C2C e-commerce services in Japan. The pipeline currently accepts logs from 5+ microservices, and the number will increase soon. - Our microservice architecture is based on the following three concepts: - 1. Split the log collection and data processing phases to keep the system simple. - 2. Use stream processing in order to achieve low latency. - 3. Don’t just accumulate raw data—support structured output that is easier to use. - Based on these concepts, we built a data pipeline in GCP. - For each microservice, we will provide a Cloud Pub/Sub “Ramp” to send logs to. Cloud Pub/Sub can have messages that contain an optional byte array in their payloads. The entire message that is ingested into the Ramp uses Cloud Dataflow streaming processing to collect them in the “RawDataHub,” a Cloud Pub/Sub topic for collection. When this happens, the PubsubMessage payload is not changed at all; the metadata necessary for subsequent processing (destination and schema information, data necessary for pipeline metrics, etc.) is provided in the PubsubMessage’s attribute map. This Dataflow job does not do processing for each service or topic—it treats all messages uniformly. - The raw data from RawDataHub is then output to two independent Cloud DataFlow streaming processes: “RawDataLake”(the infrastructure is in Google Cloud Storage, or “GCS”) and “StructuredDataHub”, another Cloud Pub/Sub topic. The StructuredDataHub has structured avro records. - The structured data from StructuredDataHub is then sent to more two independent Cloud DataFlow streaming processes: “StructuredDataLake” on GCS and “Data WareHouse” (Google BigQuery). \ No newline at end of file diff --git a/content/sessions/1323.md b/content/sessions/1323.md deleted file mode 100755 index 6f2c0fd25..000000000 --- a/content/sessions/1323.md +++ /dev/null @@ -1,10 +0,0 @@ ---- -title: "Apache Arrow + Apache Beam: A vision for cross-language, columnar data pipelines" -date: "2019-09-12T09:00:00" -track: "beam" -presenters: "Brian Hulette" ---- - - Apache Arrow and Apache Beam deserve to be together. It is a stated goal of both projects to provide mechanisms for performing data processing tasks across many languages. To accomplish this, both projects define a common binary “model” and provide multiple language implementations of that model. - However, they are approaching the overarching problem of cross-language data processing from two very different angles: Arrow is primarily concerned with moving columnar data across language boundaries with minimal overhead and providing optimized computation primitives to operate on that data, while Beam seeks to enable users to write scalable data pipelines in their language of choice and execute them on any distributed data-processing system. There is significant overlap between these two problem spaces, and we should consider using these models together for the benefit of both projects. - In this talk, I will describe both of these projects at a high-level, and present a vision for how they can be used together. I will also discuss some of the concerns around using a batched columnar data format in the Beam model, along with potential solutions. \ No newline at end of file diff --git a/content/sessions/1324.md b/content/sessions/1324.md deleted file mode 100755 index ed7d35ce6..000000000 --- a/content/sessions/1324.md +++ /dev/null @@ -1,11 +0,0 @@ ---- -title: "Stream processing for the masses with Beam, Python and Flink" -date: "2019-09-12T10:00:00" -track: "beam" -presenters: "Enrico Canzonieri" ---- - -Python is a widely used programming language that is characterized by a low barrier to entry. As many other companies in the industry, Yelp has used Python as the main programming language to implement back-end services. Unfortunately, when it comes to stream processing Python presents several challenges, including performance limitations, lack of proper multi-threading support and limited framework options. - When the use cases for more advanced stream processing started to arise, at Yelp we decided to leverage Flink and introduce a Scala/Java stack for our Data Pipeline. Over the course of two years we built our connector ecosystem in Flink and exposed to developers a very lightweight stream processing API based on Flink SQL. - While Flink SQL was quickly adopted by our developers, we soon realized that more complex use cases cannot easily be solved in Flink SQL. At the same time, the alternative engineering cost to develop and maintain a complex production application in an unfamiliar JVM based language was not convenient for our product teams. - At the beginning of the year, we started looking at Apache Beam with the goal of filling the gap between the high level Flink SQL api and the low level advanced Java Flink streaming API. In this talk I’ll cover the motivations and use cases that brought us to adopt Apache Beam, the challenges we faced integrating Beam with an existing Flink infrastructure and the strategy we followed to deploy and productionize at scale. \ No newline at end of file diff --git a/content/sessions/1325.md b/content/sessions/1325.md deleted file mode 100755 index c13779168..000000000 --- a/content/sessions/1325.md +++ /dev/null @@ -1,10 +0,0 @@ ---- -title: "Beaming Deep Learning with Ludwig" -date: "2019-09-12T13:00:00" -track: "beam" -presenters: "Suneel Marthi" ---- - -Ludwig is a code-free Deep Learning toolbox based on TensorFlow open-sourced by Uber AI Labs. Ludwig is unique in its ability to help make deep learning easier to understand for non-experts and enable faster model improvement iteration cycles for experienced machine learning developers and researchers alike. By using Ludwig, experts and researchers can simplify the prototyping process and streamline data processing so that they can focus on developing deep learning architectures rather than data wrangling. - Ludwig introduces the notion of data type-specific encoders and decoders, which results in a highly modularized and extensible architecture: each type of data supported (text, images, categories, and so on) has a specific preprocessing function. - In this talk, we’ll be looking at building Beam pipelines to programmatically create Deep Learning models with Ludwig for different input data types for both model training and inference using Beam-Python SDK. We will be showing 2 examples of training deep learning classifiers with text and images on an unbounded source and running inference on that. \ No newline at end of file diff --git a/content/sessions/1326.md b/content/sessions/1326.md deleted file mode 100755 index e354fc8df..000000000 --- a/content/sessions/1326.md +++ /dev/null @@ -1,15 +0,0 @@ ---- -title: "Pricing Lyft rides with Apache Beam - a case study in migrating from a worker-based workflow to streaming" -date: "2019-09-12T14:00:00" -track: "beam" -presenters: "Rakesh Kumar" ---- - -Ride-sharing is a two-sided marketplace; balancing supply and demand with price in real time is critical to maintaining an efficient system. Dynamic pricing creates fairness for drivers (by raising rates when there is a lot of demand) and maintains good experiences for passengers (by satisfying pick-up time SLAs). This complex system makes real-time decisions using various data sources; machine learning models; and a streaming infrastructure for low latency, reliability and scalability. In this streaming infrastructure, our system consumes a massive number of events from different sources to make these pricing decisions. - - Reacting to these events in a cron scheduler based legacy infrastructure with inherent latency becomes a challenge, especially where timely reactions are required to balance market conditions. By leveraging Apache Beam, Lyft’s Streaming Platform powers pricing by bringing together the best of two worlds: ML models in Python and Apache Flink on JVM as the streaming engine.Topics covered in this talk include: - * A brief discussion of dynamic pricing, including motivation and high-level problem formulation - * Comparison of legacy architecture and new streaming architecture - * Overview of streaming platform architecture and technology stack - * Major gains from streaming architecture - * Lessons learned \ No newline at end of file diff --git a/content/sessions/1327.md b/content/sessions/1327.md deleted file mode 100755 index 317e0ae1b..000000000 --- a/content/sessions/1327.md +++ /dev/null @@ -1,12 +0,0 @@ ---- -title: "Writing the Hazelcast Jet Runner" -date: "2019-09-12T15:30:00" -track: "beam" -presenters: "Jozsef Bartok" ---- - -Hazelcast Jet is a distributed data processing engine that threats all data as a stream. Jet is built on top of Hazelcast IMDG and thanks to this the Jet cluster can also play the role of the data source and sink. If you use Jet this way, you can achieve perfect data locality and top-of- the-class throughputs. - Jet uses cooperative multithreading (comparable to green threads), it's processors correspond to standalone single-threaded tasks that process streams of data. By not depending on OS-level thread scheduling Jet can achieve greater throughput and better saturate the CPU cores. Since the theory underlying Apache Beam has also influenced the development of Jet and there are many conceptual similarities, attempting to write a Runner based on Jet came naturally to us. When we started however the first big obstacle we hit was the relative lack of documentation. Figuring out what exactly a Runner needed to do required reverse engineering existing runners. - A second obstacle turned out to be the generic nature of the user defined functions used by Beam. While Jet's cooperative execution model allows for great performance, it requires the processing of the data to be non-blocking and this is hard to assure when running completely opaque and generic DoFns. - Yet another difficulty was presented by trying to adapt Jet's fault tolerance and processing guarantee mechanism for the Runner, a process that, even though difficult, might be more solvable than the above mentioned non-blocking requirement. - Despite these difficulties implementing an initial version of our runner turned out to be quite doable, even though experimental it's quite capable of running a was majority of the existing Beam test suits and we are proud to say that we are now listed on the Capability Matrix. We can also showcase Nexmark test results of our runner and describe possible avenues for future development. \ No newline at end of file diff --git a/content/sessions/1328.md b/content/sessions/1328.md deleted file mode 100755 index a77f809b6..000000000 --- a/content/sessions/1328.md +++ /dev/null @@ -1,9 +0,0 @@ ---- -title: "Creating Custom Streaming Events to Learn Apache Beam" -date: "2019-09-12T16:30:00" -track: "beam" -presenters: "Chengzhi Zhao" ---- - -Learning a streaming framework like Apache Beam is exciting but with 'Hello World!' examples aren't fun. There aren't a lot of free and interesting streaming data sources for beginners to play with, it's very easy to give up learning if it's boring. To keep myself learning something new, I find myself need incentives and accomplishments to continue. If you are like me need some motivation to keep learning, this talk will give you some inspiration. - In this talk, I will share my experience of learning Apache Beam by showing demos to you on how I create a streaming data with my custom 'Marvel Fights Stream Producer'. I will discuss how I went through with the Apache Beam Programming Guide and replaced the official examples with Marvel streaming data I produced. I will also talk about what I learn from creating my custom data stream producer and how that helps me learn Apache Beam better. \ No newline at end of file diff --git a/content/sessions/1329.md b/content/sessions/1329.md deleted file mode 100755 index e5fc5516d..000000000 --- a/content/sessions/1329.md +++ /dev/null @@ -1,9 +0,0 @@ ---- -title: "Production-ready stream data pipeline in Merpay, Inc" -date: "2019-09-12T17:30:00" -track: "beam" -presenters: "syucream" ---- - -We’ve started to provide our stream based data pipeline by using Google Cloud Dataflow and Apache Beam since 2018 fall. It collects event logs from microservices running on GKE, then transforms and forwards the logs to GCS and BigQuery to use for analytics, etc. As you know, implementing and operating streaming jobs are challenging. We’re encountered various issues during that time. - I’d like to share our knowledge on development and operation perspective. There are 3 topics in the development part, 1) Implementing stream jobs with using spotify/scio, 2) How to debug the jobs, especially having DynamicDestination, 3) How to load testing, to ensure our jobs stable. And topics in the next part, 1) How to deploy new jobs safely(with avoiding data loss), 2) How to monitor the jobs and surrounding systems, and misc. \ No newline at end of file diff --git a/content/sessions/1333.md b/content/sessions/1333.md deleted file mode 100755 index 56d065cf5..000000000 --- a/content/sessions/1333.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Apache@Google, an account of the success of open source practices at Google Cloud" -date: "2019-09-10T18:00:00" -track: "community" -presenters: "Kenneth Knowles" ---- - -We'll share with the Apache Community the journey that Google Cloud has had since we donated the first project to the foundation. We'll share our experiences from the commercial vendor point of view and also what we have learned while trying to embrace an "open source first" culture. \ No newline at end of file diff --git a/content/sessions/3001.md b/content/sessions/3001.md deleted file mode 100755 index 8f4d2a50c..000000000 --- a/content/sessions/3001.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "Cloudstack Locking Service" -date: "2019-09-09T15:00:00" -track: "cloudstack" -presenters: "Abhishek Kumar" ---- - -As CloudStack cannot work with any mysql clustering solution, it is time to explore a new locking service, manager and pluggable interface which would allow CloudStack DB to be HA enabled with multi-master read/write. Talk will focus on, - Need for a locking service and challenges with existing CloudStack architecture - Different possible clustering solution that can be adopted - Showcasing a PoC for future implementation with minimal changes to existing architecture using percona xtradb or any other clustering solution - Additionally, explore the idea of getting rid of mshost table, and use locking service to find about other management servers. \ No newline at end of file diff --git a/content/sessions/3002.md b/content/sessions/3002.md deleted file mode 100755 index d9481cb97..000000000 --- a/content/sessions/3002.md +++ /dev/null @@ -1,8 +0,0 @@ ---- -title: "CloudStack Monitoring - In-Depth Checks For Production Deployments" -date: "2019-09-10T17:00:00" -track: "cloudstack" -presenters: "K B Shiv Kumar" ---- - -This talk is for CloudStack administrators who're responsible for maintaining a robust and fault-free production deployment. It covers monitoring fine grained parameters of a CloudStack installation and how to measure and check not just discrete values but also routinely performed actions against their expected outcome. It also dwells upon how the same discrete values make a difference from an access perspective. Whether you want to auto-resolve common problems, provide an SLA or just keep a tab on your deployment, this talk will definitely provide some insights. \ No newline at end of file diff --git a/content/sessions/agenticcoding-1185936.md b/content/sessions/agenticcoding-1185936.md new file mode 100644 index 000000000..cbdc09da5 --- /dev/null +++ b/content/sessions/agenticcoding-1185936.md @@ -0,0 +1,20 @@ +--- +title: "Engineering BanyanDB with AI Agents: Agentic Coding in Practice" +date: "2026-08-09T13:30:00" +track: "agenticcoding" +presenters: "Youliang Huang" +stype: "Chinese Session" +room: "JingMing Hall" +--- + +This session explores how agentic coding is applied to the development of BanyanDB. Through practical examples from system design, implementation, debugging and testing, we demonstrate how engineers and AI agents can work together across different stages of software development. +Rather than focusing on code generation alone, the talk presents agentic coding as an engineering workflow in which engineers define goals, constraints, and architectural principles, while agents assist with exploration, execution, validation, and iteration. The session will also summarize practical lessons and discuss how engineering constraints and validation can make agentic coding reliable and controllable in complex systems. + +### Speakers: + + +
+ +Youliang Huang: Apache SkyWalking PMC Member + +Open Source Enthusiast. Apache SkyWalking PMC Member. Building high-performance observability, storage, and network systems. \ No newline at end of file diff --git a/content/sessions/agenticcoding-1185936.zh.md b/content/sessions/agenticcoding-1185936.zh.md new file mode 100644 index 000000000..f6f18ebc9 --- /dev/null +++ b/content/sessions/agenticcoding-1185936.zh.md @@ -0,0 +1,19 @@ +--- +title: "用 AI Agent 构建 BanyanDB：Agentic Coding 工程实践" +date: "2026-08-09T13:30:00" +track: "agenticcoding" +presenters: "Youliang Huang" +stype: "中文演讲" +room: "静明厅" +--- + +本次分享将介绍 Agentic Coding 在 BanyanDB 研发中的实际应用。通过系统设计、代码实现、问题调试和测试等实践案例，展示工程师与 AI Agent 如何在软件开发的不同环节中协同工作。 +分享不会局限于代码生成，而是将 Agentic Coding 视为一种工程协作流程：工程师负责定义目标、约束和架构原则，Agent 则协助完成探索、执行、验证和迭代。同时，分享还将总结相关实践经验，并讨论如何通过工程约束和验证机制，让 Agentic Coding 在复杂系统研发中更加可靠、可控。 + +### 讲师: + +
+ +Youliang Huang：Apache SkyWalking PMC Member + +开源爱好者。Apache SkyWalking PMC Member。致力于构建高性能的可观测性、存储与网络系统。 \ No newline at end of file diff --git a/content/sessions/agenticcoding-1207762.md b/content/sessions/agenticcoding-1207762.md new file mode 100644 index 000000000..f53c7a85f --- /dev/null +++ b/content/sessions/agenticcoding-1207762.md @@ -0,0 +1,27 @@ +--- +title: "Spec-Driven Development at Scale" +date: "2026-08-09T14:00:00" +track: "agenticcoding" +presenters: "Zhangjian He" +stype: "Chinese Session" +room: "JingMing Hall" +--- + +As AI-assisted development becomes mainstream, traditional code organization and development workflows are increasingly insufficient. Large-scale codebases, multi-repository environments, and legacy enterprise systems introduce significant challenges for applying AI effectively. + +This session introduces Spec-Driven Development (SDD) as a practical approach to scaling AI-assisted engineering. Instead of relying on ad-hoc prompts, SDD treats specifications as first-class artifacts that drive code generation, validation, and evolution. + +The talk focuses on real-world challenges and solutions in large-scale environments: + +- Codebase Structuring: designing repository and module structures that are AI-friendly and spec-aligned +- Cross-Repository Coordination: managing dependencies and changes across multiple repositories using spec-driven workflows +- Enterprise Integration: integrating SDD into existing systems, CI/CD pipelines, and development processes + +### Speakers: + + +
+ +Zhangjian He: Huawei Cloud Senior Engineer | Huawei Cloud Open Source Team Member + +Open-source enthusiast and experienced engineer. Since starting software development in 2017, I have played a key role in leading the cloud-native transformation of Huawei Cloud's IoT platform and the development of an industrial IoT platform. I serve as an Apache BookKeeper/ServiceComb PMC member, Apache Pulsar Committer, and CNCF openGemini Maintainer. I specialize in coding, software engineering, open source, and ecosystem development, driving best practices and development standards. \ No newline at end of file diff --git a/content/sessions/agenticcoding-1207762.zh.md b/content/sessions/agenticcoding-1207762.zh.md new file mode 100644 index 000000000..88595376e --- /dev/null +++ b/content/sessions/agenticcoding-1207762.zh.md @@ -0,0 +1,27 @@ +--- +title: "规模化场景下的规约驱动开发（Spec-Driven Development）" +date: "2026-08-09T14:00:00" +track: "agenticcoding" +presenters: "Zhangjian He" +stype: "中文演讲" +room: "静明厅" +--- + +随着 AI 辅助开发成为主流，传统的代码组织方式与开发工作流正日益显得力不从心。大规模代码库、多仓库（multi-repository）环境，以及老旧的企业系统，都为有效应用 AI 带来了严峻挑战。 + +本次演讲介绍规约驱动开发（Spec-Driven Development，SDD），作为规模化 AI 辅助工程的一种实践方法。SDD 不再依赖临时的（ad-hoc）prompt，而是把规约（specification）视为一等公民，用以驱动代码生成、验证与演进。 + +本次演讲聚焦大规模环境下的真实挑战与解决方案： + +- 代码库结构化：设计对 AI 友好、与规约对齐的仓库与模块结构 +- 跨仓库协同：使用规约驱动的工作流，管理多个仓库之间的依赖与变更 +- 企业集成：将 SDD 集成到现有系统、CI/CD 流水线以及开发流程中 + +### 讲师: + + +
+ +Zhangjian He：华为云高级工程师 | 华为云开源团队成员 + +开源爱好者，经验丰富的工程师。自 2017 年开始软件开发以来，我在华为云物联网平台的云原生转型以及工业物联网平台的开发中扮演了关键角色。我担任 Apache BookKeeper/ServiceComb 的 PMC 成员、Apache Pulsar Committer，以及 CNCF openGemini Maintainer。我的专长涵盖编码、软件工程、开源与生态建设，致力于推动最佳实践与开发标准。 \ No newline at end of file diff --git a/content/sessions/agenticcoding-1212968.md b/content/sessions/agenticcoding-1212968.md new file mode 100644 index 000000000..1c1f12961 --- /dev/null +++ b/content/sessions/agenticcoding-1212968.md @@ -0,0 +1,24 @@ +--- +title: "Core Design Gets Harder: Lessons from Rewriting Bub" +date: "2026-08-09T15:15:00" +track: "agenticcoding" +presenters: "卓燃 尚" +stype: "Chinese Session" +room: "JingMing Hall" +--- + +Agentic coding makes it easier than ever to generate features. But that is exactly why extensibility is not enough: we need systems that remain understandable, replaceable, and governable. + +This talk uses the rewrite of Bub as a case study in a deliberately counterintuitive design choice. At a time when AI is increasingly good at producing demos and edge features, Bub’s core is still largely designed and handwritten by humans. Bub grew out of messy multi-person group chats rather than polished single-user demos, which exposed structural problems early: once channels, skills, and memory-related capabilities accumulated, continuing to absorb them into the framework core would make the system harder to understand, configure, and maintain. + +I will share why the rewrite pushed more variation out of the framework core, kept the system small and hook-first, and treated built-ins as replaceable default plugins. I will also discuss Bub’s tape model as a way to make context more explicit, reconstructable, and reviewable for both humans and agents. The broader argument is practical: when feature generation gets cheaper, careful human design of core boundaries, dependency direction, extension points, and context becomes more important, not less. + + +### Speakers: + + +
+ +卓燃 尚: ASF Member, Senior Engineer @ OceanBase + +Data Systems · AI Infrastructure · OSS \ No newline at end of file diff --git a/content/sessions/agenticcoding-1212968.zh.md b/content/sessions/agenticcoding-1212968.zh.md new file mode 100644 index 000000000..90141ecce --- /dev/null +++ b/content/sessions/agenticcoding-1212968.zh.md @@ -0,0 +1,24 @@ +--- +title: "核心设计变得更难了：重写 Bub 的经验教训" +date: "2026-08-09T15:15:00" +track: "agenticcoding" +presenters: "卓燃 尚" +stype: "中文演讲" +room: "静明厅" +--- + +Agentic coding 让生成功能变得前所未有地容易。但这恰恰说明，仅有可扩展性已经不够：我们需要的是始终保持可理解、可替换、可治理的系统。 + +本次演讲以 Bub 的重写为案例，探讨一种刻意"反直觉"的设计选择。在 AI 越来越擅长产出 demo 和边缘功能的当下，Bub 的核心很大程度上仍由人类设计并手工编写。Bub 脱胎于杂乱的多人群聊，而非精致的单用户 demo，这早早暴露了结构性问题：一旦频道（channel）、技能（skill）以及与记忆相关的能力不断累积，继续把它们吸收进框架核心，只会让系统越来越难以理解、配置和维护。 + +我将分享为什么这次重写把更多的变化推出了框架核心，让系统保持小巧且以 hook 为先，并把内置功能当作可替换的默认插件。我还会讨论 Bub 的 tape 模型——它使上下文对人和 agent 都更明确、可重建、可审查。更宏观的论点是切实的：当功能生成的成本降低时，由人精心设计核心边界、依赖方向、扩展点和上下文，这件事会变得更加重要，而不是更不重要。 + + +### 讲师: + + +
+ +卓燃 尚：ASF Member，OceanBase 资深工程师 + +数据系统 · AI 基础设施 · 开源（OSS） \ No newline at end of file diff --git a/content/sessions/agenticcoding-1213218.md b/content/sessions/agenticcoding-1213218.md new file mode 100644 index 000000000..21e326ee4 --- /dev/null +++ b/content/sessions/agenticcoding-1213218.md @@ -0,0 +1,21 @@ +--- +title: "Agents as Testbed: Using Agents to Navigate High-Dimensional Optimization Spaces" +date: "2026-08-09T15:45:00" +track: "agenticcoding" +presenters: "Ruihang Xia" +stype: "Chinese Session" +room: "JingMing Hall" +--- + +We may have had many sparks of performance optimization, and wanna a quick PoC to verify if one idea works or not. But this procedure is very time-consuming, especially in a mature and complex system. This topic shares my practice of leveraging LLM Agent to accelerate my performance optimization tasks in real product systems by having agents do "quick tries" for me and leave the precious human decision to real critical things. And some lessons learned from it, like this method can't reduce your work but makes you even more tired :P + +### Speakers: + + +
+ +Ruihang Xia: Software Architect at Greptime Inc. + +Software Architect at Greptime +Apache DataFusion PMC, Apache Arrow Committer, Apache HoraeDB PPMC +https://github.com/waynexia \ No newline at end of file diff --git a/content/sessions/agenticcoding-1213218.zh.md b/content/sessions/agenticcoding-1213218.zh.md new file mode 100644 index 000000000..05bcd0996 --- /dev/null +++ b/content/sessions/agenticcoding-1213218.zh.md @@ -0,0 +1,21 @@ +--- +title: "把 Agent 当作试验台：用 Agent 探索高维优化空间" +date: "2026-08-09T15:45:00" +track: "agenticcoding" +presenters: "Ruihang Xia" +stype: "中文演讲" +room: "静明厅" +--- + +我们或许有过许多性能优化方面的灵感火花，想做一次快速的 PoC 来验证某个想法是否可行。但这个过程非常耗时，尤其是在一个成熟且复杂的系统中。本议题分享我的实践：利用 LLM Agent 来加速真实产品系统中的性能优化任务——让 agent 替我做"快速尝试"，而把宝贵的人类决策留给真正关键的事情。以及从中获得的一些经验教训，比如这种方法并不能减少你的工作量，反而会让你更累 :P + +### 讲师: + + +
+ +Ruihang Xia：Greptime Inc. 软件架构师 + +Greptime 软件架构师 +Apache DataFusion PMC，Apache Arrow Committer，Apache HoraeDB PPMC +https://github.com/waynexia \ No newline at end of file diff --git a/content/sessions/agenticcoding-1213385.md b/content/sessions/agenticcoding-1213385.md new file mode 100644 index 000000000..09ece4808 --- /dev/null +++ b/content/sessions/agenticcoding-1213385.md @@ -0,0 +1,42 @@ +--- +title: "Agentic Coding Unleashed From Autocomplete to Autonomous" +date: "2026-08-09T14:30:00" +track: "agenticcoding" +presenters: "Wentao Liu, Zeguo Mao, Xianxiong Luo" +stype: "Chinese Session" +room: "JingMing Hall" +--- + +Agentic Coding isn’t just about smarter autocomplete—it’s a paradigm shift where AI‑driven tools proactively understand context, make autonomous decisions, and collaborate across platforms to build, integrate, and maintain complex systems. In this talk, we’ll move beyond simple code generation to demonstrate how Cursor (the AI‑native editor), Trae (the AI‑powered testing & deployment agent), and OpenClaw (the embeddable AI assistant) work together as a cohesive, multi‑agent development squad. We’ll walk through a real‑world scenario: autonomously designing, implementing, testing, and deploying an intelligent chat‑room integration, showcasing how Agentic Coding reduces human‑to‑tool friction and elevates developers from coders to system architects. + +### Speakers: + + +
+ +Wentao Liu: Manager of omfoss.com + +Wentao Liu has been a passionate contributor to the Apache Software Foundation for many years. With over two decades of experience in the Linux and open‑source ecosystem, he has played a pivotal role in advancing the OpenMeetings project—from core development and community building to real‑world enterprise adoption. + +His contributions include: +Delivering talks and workshops about Apache OpenMeetings at major open‑source conferences such as ApacheCon, FOSSASIA, FOSDEM, and CommunityOverCode, helping to evangelize the platform’s capabilities to a global audience. +Founding a company that builds commercial video‑conference appliances based on Apache OpenMeetings, demonstrating how the project can be tailored for secure, scalable, on‑premise deployments. +Driving the integration of modern authentication (OAuth2, LDAP), virtual backgrounds, mobile applications, and clustering features that make OpenMeetings ready for enterprise use. + + +
+ +Zeguo Mao: Sophomore + +Zeguo Mao is a sophomore at Central South University of Forestry and Technology, majoring in computer science and Technology. With a strong foundation in C, C++, Java, and web technologies including SpringBoot, HTML/CSS, and JavaScript, Mao has cultivated a practical, project-driven approach to building interactive systems. His technical curiosity extends deeply into the realm of Artificial Intelligence, where he possesses a solid understanding of machine learning and deep learning algorithms. + +Mao's hands-on experience with intelligent devices, particularly the development of voice-interaction applications for Rokid smart glasses, gives him a unique perspective on human-computer collaboration. He led the "Universal Manual" project, creating a fully voice-controlled object recognition and intelligent explanation system that demonstrates the seamless integration of AR hardware with AI-driven context awareness. Additionally, as the principal investigator for the "Betel Nut Automatic Picking Robot" project, Mao was responsible for market research, feasibility analysis, and the comprehensive technical design of an autonomous system. + + +
+ +Xianxiong Luo: organizer of GitCoffee Open Source Community + +Xianxiong is a seasoned technology leader with 17+ years of hands-on experience in software development, enterprise architecture, and technical team management across E-commerce, Ride-hailing, and Open Source ecosystems. Former Product Manager for the Membership Business Line at Better Life Group and former Architect & Technical Director at Guangcai Shenzhou Group. Currently serving as the Head of Yifei Open Source and Lead Maintainer of GitCoffee Open Source Community. + +Recognized as a 2024 & 2025 Outstanding Partner of the OpenTiny Open Source Community (under Huawei Cloud). Deeply focused on bridging open-source community building with enterprise-grade AI applications. Core technical expertise spans Java Backend Architecture, Full-Stack Development, and AI Agent Implementation. \ No newline at end of file diff --git a/content/sessions/agenticcoding-1213385.zh.md b/content/sessions/agenticcoding-1213385.zh.md new file mode 100644 index 000000000..df5adfe1a --- /dev/null +++ b/content/sessions/agenticcoding-1213385.zh.md @@ -0,0 +1,42 @@ +--- +title: "Agentic Coding 全面释放：从自动补全到自主执行" +date: "2026-08-09T14:30:00" +track: "agenticcoding" +presenters: "Wentao Liu, Zeguo Mao, Xianxiong Luo" +stype: "中文演讲" +room: "静明厅" +--- + +Agentic Coding 不仅仅是更聪明的自动补全——它是一次范式转变：AI 驱动的工具会主动理解上下文、自主做出决策，并跨平台协作，以构建、集成和维护复杂系统。在本次演讲中，我们将超越简单的代码生成，演示 Cursor（AI 原生编辑器）、Trae（AI 驱动的测试与部署 agent）以及 OpenClaw（可嵌入的 AI 助手）如何作为一个紧密协作的多 agent 开发小队协同工作。我们将走完一个真实场景：自主地设计、实现、测试并部署一个智能聊天室集成，展示 Agentic Coding 如何减少人与工具之间的摩擦，并让开发者从"码农"跃升为系统架构师。 + +### 讲师: + + +
+ +Wentao Liu：omfoss.com 负责人 + +Wentao Liu 多年来一直是 Apache 软件基金会的热情贡献者。他在 Linux 与开源生态拥有超过二十年的经验，在推动 OpenMeetings 项目的发展中发挥了关键作用——从核心开发、社区建设，到真实的企业落地。 + +他的贡献包括： +在 ApacheCon、FOSSASIA、FOSDEM、CommunityOverCode 等主要开源会议上发表关于 Apache OpenMeetings 的演讲和工作坊，帮助向全球受众推广该平台的能力。 +创办了一家公司，基于 Apache OpenMeetings 打造商用视频会议设备，展示了该项目如何被定制为安全、可扩展的本地化（on-premise）部署。 +推动现代认证（OAuth2、LDAP）、虚拟背景、移动应用以及集群等特性的集成，使 OpenMeetings 真正具备企业级使用条件。 + + +
+ +Zeguo Mao：大二学生 + +Zeguo Mao 是中南林业科技大学计算机科学与技术专业的大二学生。他在 C、C++、Java 以及包括 SpringBoot、HTML/CSS、JavaScript 在内的 Web 技术方面基础扎实，形成了一种以实践、项目驱动的方式构建交互系统的方法。他的技术好奇心还深入人工智能领域，对机器学习和深度学习算法有着扎实的理解。 + +Mao 在智能设备方面的实战经验——尤其是为 Rokid 智能眼镜开发语音交互应用——赋予了他对人机协作的独特视角。他牵头了"通用说明书"项目，打造了一套全语音控制的目标识别与智能讲解系统，展示了 AR 硬件与 AI 驱动的上下文感知之间的无缝融合。此外，作为"槟榔自动采摘机器人"项目的负责人，Mao 负责市场调研、可行性分析，以及一整套自主系统的综合技术设计。 + + +
+ +Xianxiong Luo：GitCoffee 开源社区发起人 + +Xianxiong 是一位资深的技术领导者，在软件开发、企业架构和技术团队管理方面拥有 17 年以上的实战经验，横跨电商、网约车和开源生态。他曾任步步高集团（Better Life Group）会员业务线产品经理，以及光彩神州集团（Guangcai Shenzhou Group）的架构师与技术总监。目前担任益飞开源（Yifei Open Source）负责人，以及 GitCoffee 开源社区的首席维护者。 + +他曾被评为 OpenTiny 开源社区（隶属华为云）2024 及 2025 年度优秀合作伙伴。他深度聚焦于把开源社区建设与企业级 AI 应用连接起来。核心技术专长涵盖 Java 后端架构、全栈开发以及 AI Agent 落地实现。 \ No newline at end of file diff --git a/content/sessions/aiinfra-1208628.md b/content/sessions/aiinfra-1208628.md new file mode 100644 index 000000000..3c3743894 --- /dev/null +++ b/content/sessions/aiinfra-1208628.md @@ -0,0 +1,32 @@ +--- +title: "Building a Multimodal AI Lakehouse with Apache Gravitino in China Mobile Wutong Data Platform" +date: "2026-08-07T15:00:00" +track: "aiinfra" +presenters: "Xiaojing Fang, Xintong Jiang" +stype: "Chinese Session" +room: "YuanMing Hall" +--- + +As AI workloads become a core part of modern data platforms, enterprises need to manage not only structured tables, but also unstructured data, vector data, AI models, and AI functions in a unified way. Traditional lakehouse architectures are often centered on tabular data only, which makes it difficult to build a consistent metadata and control plane across heterogeneous data and AI assets. + +In this session, we will share how China Mobile is building a multimodal AI lakehouse in the Wutong Data Platform with Apache Gravitino. In this architecture, Gravitino serves as the unified metadata and control plane to manage structured and unstructured data together with AI models and AI functions. This enables consistent governance, discovery, and management across different types of assets in the platform. + +We will also discuss how the platform combines Apache Iceberg and Lance to support unified analytics across tabular and multimodal data. In addition to introducing the relevant capabilities of Apache Gravitino, this talk will present practical experience from China Mobile’s production platform, including architecture design, metadata organization, integration patterns, governance considerations, and lessons learned from real-world deployment. + +This session is intended for data platform engineers, lakehouse architects, and AI infrastructure practitioners who are exploring open architectures for multimodal data and AI asset management. + +### Speakers: + + +
+ +Xiaojing Fang: Apache Gravitino Committer + +Apache Gravitino PPMC, architect at China Mobile, focusing on data and AI infrastructure. + + +
+ +Xintong Jiang: Software R&D Engineer at ChinaMobile + +Boasting more than a decade of expertise in big data and AI, spearheads the development of PB-scale multimodal data lakes at China Mobile Digital Intelligence Business Unit.Core competencies lie in distributed computing and hybrid retrieval. \ No newline at end of file diff --git a/content/sessions/aiinfra-1208628.zh.md b/content/sessions/aiinfra-1208628.zh.md new file mode 100644 index 000000000..8a6b37a51 --- /dev/null +++ b/content/sessions/aiinfra-1208628.zh.md @@ -0,0 +1,32 @@ +--- +title: "基于 Apache Gravitino 在中国移动梧桐数据平台构建多模态 AI 湖仓" +date: "2026-08-07T15:00:00" +track: "aiinfra" +presenters: "Xiaojing Fang, Xintong Jiang" +stype: "中文演讲" +room: "圆明厅" +--- + +随着 AI 工作负载成为现代数据平台的核心组成部分，企业不仅需要管理结构化表，还需要以统一的方式管理非结构化数据、向量数据、AI 模型和 AI 函数。传统的湖仓架构往往只以表格数据为中心，这使得在异构数据与 AI 资产之间构建一致的元数据和控制面变得困难。 + +在本次演讲中，我们将分享中国移动如何基于 Apache Gravitino 在梧桐数据平台构建多模态 AI 湖仓。在这一架构中，Gravitino 作为统一的元数据和控制面，把结构化、非结构化数据与 AI 模型、AI 函数一并纳管。这使得平台中不同类型的资产能够获得一致的治理、发现与管理能力。 + +我们还将讨论该平台如何结合 Apache Iceberg 和 Lance，以支撑对表格数据和多模态数据的统一分析。除了介绍 Apache Gravitino 的相关能力外，本次演讲还将呈现来自中国移动生产平台的实战经验，涵盖架构设计、元数据组织、集成模式、治理考量，以及真实部署中总结的经验教训。 + +本次演讲面向数据平台工程师、湖仓架构师以及 AI 基础设施从业者，适合那些正在探索多模态数据与 AI 资产管理开放架构的人士。 + +### 讲师: + + +
+ +Xiaojing Fang：Apache Gravitino Committer + +Apache Gravitino PPMC，中国移动架构师，专注于数据与 AI 基础设施。 + + +
+ +Xintong Jiang：中国移动软件研发工程师 + +拥有十余年大数据与 AI 领域的专业经验，在中国移动数智（Digital Intelligence）业务单元主导 PB 级多模态数据湖的开发。核心能力在于分布式计算与混合检索（hybrid retrieval）。 \ No newline at end of file diff --git a/content/sessions/aiinfra-1212172.md b/content/sessions/aiinfra-1212172.md new file mode 100644 index 000000000..eb412d90f --- /dev/null +++ b/content/sessions/aiinfra-1212172.md @@ -0,0 +1,25 @@ +--- +title: "Open Chain AI SBOM" +date: "2026-08-07T16:15:00" +track: "aiinfra" +presenters: "Meixia Wang" +stype: "Chinese Session" +room: "YuanMing Hall" +--- + +An organization shall establish AI governance frameworks, policies, and practices to ensure that AI systems are developed, deployed, and managed responsibly. Governance shall emphasize compliance with applicable and emerging AI laws and regulations (e.g., the EU AI Act), while also addressing ethical considerations, risk management, and transparency. +Based on the above principles, OpenChain has developed an AI SBOM Self-Certification program to help organizations reduce AI-related risks from an implementation perspective. This approach aligns with and supports the implementation of ISO/IEC 42001. + + +### Speakers: + + +
+ +Meixia Wang: Linux Foundation, Executive Director of Open Chain Project + +Mary Wang is the Executive Director of the OpenChain Foundation, part of the Linux Foundation. + +Prior to this role, she served as Director of the Open Source Ecosystem at Volvo Cars, where she led key initiatives including launching open source projects, building the supporting infrastructure for an open source ecosystem, and establishing the company’s Open Source Program Office (OSPO). + +Before joining Volvo Cars, Mary spent nine years at Ericsson. During that time, she worked as a DevOps engineer, developing and implementing CI/CD pipelines, and later as a Technical Product Manager (TPM), taking end-to-end ownership of products. \ No newline at end of file diff --git a/content/sessions/aiinfra-1212172.zh.md b/content/sessions/aiinfra-1212172.zh.md new file mode 100644 index 000000000..a75ee9a7d --- /dev/null +++ b/content/sessions/aiinfra-1212172.zh.md @@ -0,0 +1,25 @@ +--- +title: "OpenChain AI SBOM" +date: "2026-08-07T16:15:00" +track: "aiinfra" +presenters: "Meixia Wang" +stype: "中文演讲" +room: "圆明厅" +--- + +组织应当建立 AI 治理框架、政策和实践，以确保 AI 系统的开发、部署和管理是负责任的。治理应当强调遵守现行及新兴的 AI 法律法规（例如欧盟《AI 法案》），同时兼顾伦理考量、风险管理与透明度。 +基于上述原则，OpenChain 推出了一项 AI SBOM 自认证（Self-Certification）计划，帮助组织从落地实施的角度降低与 AI 相关的风险。这一方法与 ISO/IEC 42001 的实施相契合，并对其形成支撑。 + + +### 讲师: + + +
+ +Meixia Wang：Linux Foundation，OpenChain 项目执行董事 + +Mary Wang 是 Linux Foundation 旗下 OpenChain 基金会的执行董事。 + +在担任此职务之前，她曾任沃尔沃汽车（Volvo Cars）开源生态总监，主导了多项关键举措，包括发起开源项目、为开源生态建设配套基础设施，以及在公司内部建立开源项目办公室（OSPO）。 + +在加入沃尔沃汽车之前，Mary 曾在爱立信（Ericsson）工作九年。期间，她先是担任 DevOps 工程师，开发并实施 CI/CD 流水线；后来担任技术产品经理（TPM），对产品进行端到端的负责。 \ No newline at end of file diff --git a/content/sessions/aiinfra-1212762.md b/content/sessions/aiinfra-1212762.md new file mode 100644 index 000000000..7baf801fa --- /dev/null +++ b/content/sessions/aiinfra-1212762.md @@ -0,0 +1,28 @@ +--- +title: "Cost Effective Agent Observability and Evaluation built on Apache Doris" +date: "2026-08-07T15:45:00" +track: "aiinfra" +presenters: "Kang Xiao" +stype: "Chinese Session" +room: "YuanMing Hall" +--- + +In the age of AI, the goals of observability extend far beyond mere system stability and high performance. Given the inherent uncertainty of Large Language Models (LLMs) and AI Agents, detailed tracking, evaluation, and optimization have become more critical than ever. + +Litefuse is an Agent observability and evaluation platform built on Apache Doris. It supports evaluation-driven development, helping developers build and deliver more reliable AI Agents. + +Presentation Highlights: +1. The importance of Agent Observability and Evaluation. +2. Key challenges in Agent Observability and Evaluation. +3. Building Agent Observability and Evaluation based on Apache Doris. + +### Speakers: + + +
+ +Kang Xiao: SelectDB, Co-Founder + +SelectDB, Co-Founder +360 and QianXin, Director +Baidu, Senior Engineer \ No newline at end of file diff --git a/content/sessions/aiinfra-1212762.zh.md b/content/sessions/aiinfra-1212762.zh.md new file mode 100644 index 000000000..2e08f3a09 --- /dev/null +++ b/content/sessions/aiinfra-1212762.zh.md @@ -0,0 +1,28 @@ +--- +title: "基于 Apache Doris 构建高性价比的 Agent 可观测性与评估" +date: "2026-08-07T15:45:00" +track: "aiinfra" +presenters: "Kang Xiao" +stype: "中文演讲" +room: "圆明厅" +--- + +在 AI 时代，可观测性的目标已远远超出单纯的系统稳定性和高性能。鉴于大语言模型（LLM）和 AI Agent 固有的不确定性，细致的追踪、评估与优化变得前所未有地关键。 + +Litefuse 是一个基于 Apache Doris 构建的 Agent 可观测性与评估平台。它支持评估驱动开发（evaluation-driven development），帮助开发者构建并交付更可靠的 AI Agent。 + +演讲亮点： +1. Agent 可观测性与评估的重要性。 +2. Agent 可观测性与评估的关键挑战。 +3. 基于 Apache Doris 构建 Agent 可观测性与评估。 + +### 讲师: + + +
+ +Kang Xiao：SelectDB 联合创始人 + +SelectDB 联合创始人 +360 与奇安信（QianXin）总监 +百度高级工程师 \ No newline at end of file diff --git a/content/sessions/aiinfra-1214534.md b/content/sessions/aiinfra-1214534.md new file mode 100644 index 000000000..f82138d27 --- /dev/null +++ b/content/sessions/aiinfra-1214534.md @@ -0,0 +1,28 @@ +--- +title: "Building an AI-Native Metadata Service on Tencent Cloud with Apache Gravitino" +date: "2026-08-07T14:30:00" +track: "aiinfra" +presenters: "Jinsong Zhou, Jie Zan" +stype: "Chinese Session" +room: "YuanMing Hall" +--- + +2026 is the year AI agents went from demos to production. Every agent — whether querying data, training models, or orchestrating pipelines — hits the same wall: "Where is the data, and do I have permission?" The metadata catalog is no longer a passive registry; it is the contextual brain agents depend on to discover, reason, and act. + +This talk shares how we built TCCatalog on Apache Gravitino to serve this new role at Tencent Cloud scale. We cover cloud-native architecture, extensions for AI-native workloads (Iceberg + Lance + Model Registry + MCP), and hard-won production lessons — including real incident stories, performance data, and engineering trade-offs. + +### Speakers: + + +
+ +Jinsong Zhou: Staff software engineer at Tencent, PPMC member of Apache Amoro (incubating) + +Jin Song Zhou, a staff software engineer at Tencent, brings 10 years of expertise in the fields of databases and big data. Currently leading the implementation of a lakehouse architecture based on Apache Iceberg at Tencent Cloud, also serving as a PPMC member for Apache Amoro (incubating), actively fostering community development. + + +
+ +Jie Zan: Staff Software Engineer at Tencent Cloud. + +Jie Zan is a Staff Software Engineer at Tencent Cloud, working in the Big Data Infrastructure Platform team. He focuses on Data + AI technologies, including AI agents, metadata integration, and cloud-native data infrastructure. His work includes building metadata services based on Apache Gravitino and developing cloud-native data platforms such as TCBase BaaS to enable unified governance for data and AI workloads. \ No newline at end of file diff --git a/content/sessions/aiinfra-1214534.zh.md b/content/sessions/aiinfra-1214534.zh.md new file mode 100644 index 000000000..8407a5903 --- /dev/null +++ b/content/sessions/aiinfra-1214534.zh.md @@ -0,0 +1,28 @@ +--- +title: "基于 Apache Gravitino 在腾讯云构建 AI 原生元数据服务" +date: "2026-08-07T14:30:00" +track: "aiinfra" +presenters: "Jinsong Zhou, Jie Zan" +stype: "中文演讲" +room: "圆明厅" +--- + +2026 年是 AI agent 从 demo 走向生产的一年。每一个 agent——无论它是在查询数据、训练模型，还是编排管道——都会撞上同一堵墙："数据在哪里？我有没有权限？" 元数据 catalog 不再只是一个被动的登记簿；它是 agent 用来发现、推理和行动的"上下文大脑"。 + +本次演讲分享我们如何在 Apache Gravitino 之上构建 TCCatalog，使其在腾讯云的规模下胜任这一新角色。我们将介绍云原生架构、面向 AI 原生工作负载的扩展（Iceberg + Lance + Model Registry + MCP），以及来之不易的生产经验——包括真实故障故事、性能数据与工程权衡。 + +### 讲师: + + +
+ +Jinsong Zhou：腾讯专家软件工程师，Apache Amoro（孵化中）PPMC 成员 + +Jinsong Zhou 是腾讯的 Staff 软件工程师，在数据库和大数据领域拥有 10 年的专业经验。他目前在腾讯云牵头基于 Apache Iceberg 的湖仓架构落地，同时担任 Apache Amoro（孵化中）的 PPMC 成员，积极推动社区发展。 + + +
+ +Jie Zan: 腾讯专家软件工程师 + +昝杰（Jie Zan）现任腾讯云专家软件工程师，隶属于大数据基础平台团队。他专注于数据 + AI 技术领域，研究方向涵盖AI智能体、元数据集成及云原生数据基础设施。目前主要负责基于 Apache Gravitino 构建元数据服务，同时牵头研发 TCBase BaaS 等云原生数据平台，实现数据与 AI 工作负载的统一治理。 \ No newline at end of file diff --git a/content/sessions/aiinfra-1214790.md b/content/sessions/aiinfra-1214790.md new file mode 100644 index 000000000..b52bb140b --- /dev/null +++ b/content/sessions/aiinfra-1214790.md @@ -0,0 +1,19 @@ +--- +title: "Apache APISIX: From API Gateway to AI Gateway" +date: "2026-08-07T14:00:00" +track: "aiinfra" +presenters: "Qi Guo" +stype: "Chinese Session" +room: "YuanMing Hall" +--- + +Apache APISIX is a cloud-native API gateway that is now gradually evolving into an AI gateway as well. This talk will provide an overview of this evolution, along with some of the core features of the AI gateway and the ideas behind its implementation. + +### Speakers: + + +
+ +Qi Guo: Committer of Apache APISIX + +Committer of Apache APISIX \ No newline at end of file diff --git a/content/sessions/aiinfra-1214790.zh.md b/content/sessions/aiinfra-1214790.zh.md new file mode 100644 index 000000000..be9ea2a79 --- /dev/null +++ b/content/sessions/aiinfra-1214790.zh.md @@ -0,0 +1,19 @@ +--- +title: "Apache APISIX 从 API 网关到 AI 网关" +date: "2026-08-07T14:00:00" +track: "aiinfra" +presenters: "郭奇" +stype: "中文演讲" +room: "圆明厅" +--- + +Apache APISIX 是一个云原生 API 网关，现在也逐渐变为一个 AI 网关，这个分享中会大概介绍这个演进过程，以及 AI 网关的一些核心功能和实现思路。 + +### 讲师: + + +
+ +郭奇：Apache APISIX Committer + +Apache APISIX Committer \ No newline at end of file diff --git a/content/sessions/aiinfra-1214842.md b/content/sessions/aiinfra-1214842.md new file mode 100644 index 000000000..b482a6cd5 --- /dev/null +++ b/content/sessions/aiinfra-1214842.md @@ -0,0 +1,21 @@ +--- +title: "OpenClaw Security Governance: From Disorder to Order" +date: "2026-08-07T16:45:00" +track: "aiinfra" +presenters: "Leo (Zhiquan) Yu" +stype: "Chinese Session" +room: "YuanMing Hall" +--- + +Through real-world cases, it reveals the systemic security risks that cannot be ignored behind the explosive growth of autonomous AI agents (represented by OpenClaw), attributing their root cause to the "fatal triangle" where "functionality and security cannot coexist." Seizing this opportunity, it showcases its comprehensive product ecosystem—spanning governance philosophy, core technology (metacognition), and covering assessment, defense, training, and hardware across the entire chain—aiming to provide enterprises with a one-stop AI agent security governance solution. The ultimate goal is to lead AI from its current state of "disorder" toward a new ecosystem defined as "First Order." + +### Speakers: + + +
+ +Leo (Zhiquan) Yu: Head of AI Ecosystem and Market at Shouxu Intelligence. + +AIFun (AI Industry Exchange and Cooperation Community) Co-founder; Head of AI Ecosystem and Market at Shouxu Intelligence. With years of experience in software R&D and management, as well as technical community operations. +Previously served as PMO and Executive Committee Member of the Technology Committee at Lianshang Network (WiFi Master Key), Technical Director at Mingdao Cloud (Open Platform and App Market), Head of Haiding Network's Headin Technical Community and Event Platform, and R&D Manager at Corel. +Social roles include Executive Committee Member of the CCF TF Engineer Culture and CTO Club Shanghai for the China Computer Federation (CCF), member of the Suzhou Computer Federation (SZCF) and Shanghai Open Source Association (SHOpen), and involvement in Agile Community / Agile Tour. Deeply engaged in developer ecosystems such as Google GDG, Amazon Web Services UG, and Huawei HDG, serving as a core organizer for related technical communities. \ No newline at end of file diff --git a/content/sessions/aiinfra-1214842.zh.md b/content/sessions/aiinfra-1214842.zh.md new file mode 100644 index 000000000..aa3409a21 --- /dev/null +++ b/content/sessions/aiinfra-1214842.zh.md @@ -0,0 +1,21 @@ +--- +title: "OpenClaw 安全治理：从无序到有序" +date: "2026-08-07T16:45:00" +track: "aiinfra" +presenters: "Leo (Zhiquan) Yu" +stype: "中文演讲" +room: "圆明厅" +--- + +通过真实案例，本演讲揭示了以 OpenClaw 为代表的自主 AI agent 爆发式增长背后，那些不容忽视的系统性安全风险，并将其根因归结为"功能性与安全性无法兼得"的"致命三角"。以此为契机，展示了其完整的产品生态——涵盖治理理念、核心技术（元认知，metacognition），并覆盖评估、防御、训练与硬件的整条链路——旨在为企业提供一站式的 AI agent 安全治理方案。最终目标，是带领 AI 从当下的"无序"状态，走向一个被定义为"第一秩序（First Order）"的新生态。 + +### 讲师: + + +
+ +Leo (Zhiquan) Yu：守序智能（Shouxu Intelligence）AI 生态与市场负责人 + +AIFun（AI 产业交流与合作社区）联合创始人；守序智能（Shouxu Intelligence）AI 生态与市场负责人。在软件研发与管理、技术社区运营方面拥有多年经验。 +此前曾任连尚网络（WiFi 万能钥匙）PMO 及技术委员会执委，明道云（开放平台与应用市场）技术总监，海丁网络 Headin 技术社区与活动平台负责人，以及 Corel 的研发经理。 +社会职务包括中国计算机学会（CCF）TF 工程师文化执委与 CTO 俱乐部（上海），苏州市计算机学会（SZCF）与上海开源协会（SHOpen）成员，并参与敏捷社区 / Agile Tour。深度参与 Google GDG、Amazon Web Services UG、华为 HDG 等开发者生态，担任相关技术社区的核心组织者。 \ No newline at end of file diff --git a/content/sessions/community-1196034.md b/content/sessions/community-1196034.md new file mode 100644 index 000000000..d68a28262 --- /dev/null +++ b/content/sessions/community-1196034.md @@ -0,0 +1,25 @@ +--- +title: "Open Source Community & Developer Ecosystem AMA: How Communities Succeed Beyond Code" +date: "2026-08-07T16:00:00" +track: "community" +presenters: "Richard Lin" +stype: "Chinese Session" +room: "Mtn YuQuan Hall" +--- +The success of an open source project is never determined by code alone. Why should companies continue investing in open source and developer ecosystems? Do stars, downloads, and signups truly reflect community health? How can users become contributors and maintainers? How can a vendor-led project earn community trust while building a sustainable business? And how do we address maintainer burnout, career ceilings, and leadership succession? + +This session is not a conventional lecture. It is a small, interactive AMA where participants are encouraged to ask questions, interrupt, challenge assumptions, and share real problems from their projects, communities, and companies. + +Drawing on two books he co-translated into Chinese—*Open Source Projects: Beyond Code* and *Developer Relations: How to Build and Grow a Successful Developer Program*—Richard Lin will explore five challenges shared by open source communities and developer programs: demonstrating business value, measuring community health, improving the developer journey, balancing monetization with trust, and sustaining the people behind the ecosystem. + +There are no universal answers, but together we can identify practices that fit the realities of different projects and organizations. + +### Speakers: + +
+ +Richard Lin VP of Ecosystem at Datastrato; Co-founder and 2026 Chair of KAIYUANSHE + +Richard Lin is VP of Ecosystem at Datastrato, the company behind Apache Gravitino, and co-founder and 2026 Chair of KAIYUANSHE. He brings nearly two decades of experience building open source communities and developer ecosystems across Taiwan, mainland China, and Silicon Valley. + +Previously, Richard led open source and global developer ecosystem initiatives at 01.AI and the AI developer ecosystem at Huawei Cloud. He is also a co-translator of the Chinese editions of *Open Source Projects: Beyond Code* and *Developer Relations: How to Build and Grow a Successful Developer Program*. His work focuses on open source governance, developer experience, community growth, commercial ecosystems, and the long-term sustainability of the people behind them. \ No newline at end of file diff --git a/content/sessions/community-1196034.zh.md b/content/sessions/community-1196034.zh.md new file mode 100644 index 000000000..86f43e41c --- /dev/null +++ b/content/sessions/community-1196034.zh.md @@ -0,0 +1,30 @@ +--- +title: "开源社区运营与开发者生态 AMA：从商业价值、社区健康到人的可持续性" +date: "2026-08-07T16:00:00" +track: "community" +presenters: "Richard Lin" +stype: "中文演讲" +room: "玉泉山会议室" +--- + +开源项目的成功，从来不只取决于代码。 + +开源和开发者生态不能直接带来收入，企业为什么还要持续投入？Star、下载量和注册数，真的能够反映社区健康吗？如何帮助用户成长为贡献者和维护者？厂商主导的开源项目怎样在商业化过程中赢得信任？面对维护者倦怠、职业天花板和领导者离开，社区又该如何实现人的可持续发展？ + +本场活动不是一场单向演讲，而是一次小范围、开放式的 AMA。讲者将结合其翻译的《开源项目成功之道》与《开发者关系：方法与实践》两本书，从开源项目和企业开发者生态两个视角，围绕五个共性问题展开讨论。 + +欢迎参与者带着自己在项目、社区和企业中的真实困境来到现场，随时提问、打断、补充或提出不同意见。这里没有放之四海而皆准的标准答案，只有来自实践的经验、失败与共同探索。 + +### 讲师: + +
+ +林旅强（Richard Lin） Datastrato 生态副总裁；开源社联合创始人、2026 届理事长 + +讲师简介： + +林旅强现任 Datastrato 生态副总裁，负责 Apache Gravitino 背后团队的开源生态工作；同时担任开源社联合创始人及 2026 届理事长。 + +他拥有近 20 年开源社区与开发者生态实践经验，工作经历横跨台湾、中国大陆与硅谷。此前曾任零一万物（01.AI）开源与全球开发者生态负责人，以及华为云 AI 开发者生态负责人。 + +他也是《开源项目成功之道》和《开发者关系：方法与实践》中文版合译者，长期关注开源治理、贡献者成长、开发者体验、社区商业化与生态可持续发展。 \ No newline at end of file diff --git a/content/sessions/community-1199603.md b/content/sessions/community-1199603.md new file mode 100644 index 000000000..e2edabf8c --- /dev/null +++ b/content/sessions/community-1199603.md @@ -0,0 +1,31 @@ +--- +title: "From Legacy Governance to Community Leadership: Building a Network-Focused OSPO" +date: "2026-08-08T14:00:00" +track: "community" +presenters: "Nikola Babadzhanov" +stype: "English Session" +room: "Mtn YuQuan Hall" +--- + +How can large, established organizations move beyond internal processes and become influential leaders in the open source ecosystem? The key lies in transforming legacy governance into a springboard for external engagement. This talk shares the journey of how the Bosch Open Source Program Office (OSPO) was built on top of existing structures to foster a thriving culture of external contribution and knowledge sharing. + +We will explore practical strategies for: + +- Building a robust external network and positioning your organization as a valuable partner in the OSS community. + +- Sharing expertise and best practices to establish thought leadership. + +- Empowering and scaling contributions to critical open source projects. + +Drawing heavily on our experience at Bosch, this session provides a blueprint for how you can evolve your own legacy OSS processes from a gatekeeper into a powerful enabler for community participation and leadership. + +### Speakers: + + +
+ +Nikola Babadzhanov: Open Source Consultant + +Nikola Babadzhanov joined Bosch in 2022. He is an Open Source Consultant and a member of the Open Source Program Office. He provides Bosch-internal Open Source Services as a member of the Open Source Consulting team. + +Nikola is a member of the Center of Excellence Open Source and Inner Source at Bosch; active member of the OpenChain SBOM work group, AI work group and Tooling work group. Nikola is OpenChain Ambassador and committer on the Eclipse Apoapsis project. \ No newline at end of file diff --git a/content/sessions/community-1199603.zh.md b/content/sessions/community-1199603.zh.md new file mode 100644 index 000000000..b5d6f8d97 --- /dev/null +++ b/content/sessions/community-1199603.zh.md @@ -0,0 +1,31 @@ +--- +title: "从遗留治理到社区领导力：构建以网络为重的 OSPO" +date: "2026-08-08T14:00:00" +track: "community" +presenters: "Nikola Babadzhanov" +stype: "英文演讲" +room: "玉泉山会议室" +--- + +大型且历史悠久的组织如何才能超越内部流程，成为开源生态中有影响力的领导者？关键在于把既有的（legacy）治理转化为对外参与的跳板。本次演讲分享 Bosch 开源项目办公室（OSPO）是如何在既有组织结构之上建立起来的，从而培育出繁荣的外部贡献与知识共享文化。 + +我们将探讨以下方面的实用策略： + +- 构建稳健的外部网络，让你的组织成为 OSS 社区中有价值的合作伙伴。 + +- 分享专业经验与最佳实践，以确立思想领导力。 + +- 赋能并扩大对关键开源项目的贡献。 + +本次演讲大量借鉴我们在 Bosch 的经验，为你提供一份蓝图：如何把自己既有的 OSS 流程，从一道"把关人"，演进为推动社区参与与领导力的强大赋能者。 + +### 讲师: + + +
+ +Nikola Babadzhanov：开源顾问 + +Nikola Babadzhanov 于 2022 年加入 Bosch。他是一名开源顾问，也是开源项目办公室（OSPO）的成员。作为开源咨询团队的成员，他为 Bosch 内部提供开源服务。 + +Nikola 是 Bosch 卓越中心（Center of Excellence）开源与 Inner Source 的成员；OpenChain SBOM 工作组、AI 工作组和 Tooling 工作组的活跃成员。Nikola 是 OpenChain Ambassador，也是 Eclipse Apoapsis 项目的 committer。 \ No newline at end of file diff --git a/content/sessions/community-1208527.md b/content/sessions/community-1208527.md new file mode 100644 index 000000000..b541ca97a --- /dev/null +++ b/content/sessions/community-1208527.md @@ -0,0 +1,21 @@ +--- +title: "The Open Source Way in the AI Era: Exploring the Diversity of Apache and CNCF Governance Styles" +date: "2026-08-07T14:00:00" +track: "community" +presenters: "Fengjun Lyu" +stype: "Chinese Session" +room: "Mtn YuQuan Hall" +--- + +In the AI era, where technology iterates exponentially, the philosophy of "Community Over Code" is more critical than ever. As an active Apache PMC member, a CNCF project Maintainer, and an enterprise OSPO practitioner, I will share the diverse governance styles and successful practices of these two top-tier open-source foundations. + +This session is not about comparing which is better; rather, it focuses on the operational wisdom behind different governance models. We will dive deep into their governance structures, contributor incentive mechanisms, and community event organization. Combined with real-world success stories, we will discuss how open-source communities can flexibly adapt to the new trends and challenges in the AI era, and how to harness the power of the community to drive continuous innovation. Regardless of your technical ecosystem, you will walk away with practical guidelines for participating in, understanding, and building thriving open-source communities. + +### Speakers: + + +
+ +Fengjun Lyu: Ant Group, Technical Evangelist + +I'm Mingcheng (明城). Originally from Ningbo, I am currently based in Hangzhou, China. I work as an open-source evangelist at Ant Group and serve on the Project Management Committee (PMC) at the Apache Software Foundation. I actively maintain several open-source projects, including InclusionAI, Dragonfly, and Apache Geaflow. \ No newline at end of file diff --git a/content/sessions/community-1208527.zh.md b/content/sessions/community-1208527.zh.md new file mode 100644 index 000000000..086198812 --- /dev/null +++ b/content/sessions/community-1208527.zh.md @@ -0,0 +1,21 @@ +--- +title: "AI 时代的开源之道：探索 Apache 与 CNCF 治理风格的多样性" +date: "2026-08-07T14:00:00" +track: "community" +presenters: "Fengjun Lyu" +stype: "中文演讲" +room: "玉泉山会议室" +--- + +在技术呈指数级迭代的 AI 时代，"Community Over Code（社区重于代码）"的理念比以往任何时候都更为关键。作为一名活跃的 Apache PMC 成员、CNCF 项目 Maintainer，以及企业 OSPO 实践者，我将分享这两大顶级开源基金会多样化的治理风格与成功实践。 + +本次演讲不是要比出孰优孰劣，而是聚焦于不同治理模式背后的运营智慧。我们将深入剖析它们的治理结构、贡献者激励机制以及社区活动组织。结合真实的成功案例，讨论开源社区如何灵活应对 AI 时代的新趋势与新挑战，以及如何借助社区的力量驱动持续创新。无论你身处何种技术生态，都能带走参与、理解并建设繁荣开源社区的实用指南。 + +### 讲师: + + +
+ +Fengjun Lyu：蚂蚁集团，技术布道师 + +我是 Mingcheng（明城）。来自宁波，目前定居中国杭州。我在蚂蚁集团担任开源布道师，并任职于 Apache 软件基金会的项目管理委员会（PMC）。我积极维护多个开源项目，包括 InclusionAI、Dragonfly 和 Apache Geaflow。 \ No newline at end of file diff --git a/content/sessions/community-1210509.md b/content/sessions/community-1210509.md new file mode 100644 index 000000000..86f1fd7cf --- /dev/null +++ b/content/sessions/community-1210509.md @@ -0,0 +1,30 @@ +--- +title: "To Ban or Not to Ban? How Open Source Projects Govern GenAI Contributions" +date: "2026-08-07T14:30:00" +track: "community" +presenters: "Wenhao Yang, Runzhi He" +stype: "Chinese Session" +room: "Mtn YuQuan Hall" +--- + +Generative AI now writes code, opens issues, files pull requests, reviews changes, and even reports vulnerabilities in open source projects. But cheaper generation does not mean cheaper review. Under the growing maintenance burden, projects have started writing GenAI-specific rules — in CONTRIBUTING files, security policies, PR templates, and AGENTS.md — ranging from mandatory disclosure all the way to total bans. + +To understand what is actually emerging, we read the AI rules of 67 of the most visible open source projects, along with the incidents and debates behind them. The picture goes far beyond ban-or-not. Maintainers' worries cluster around seven pressure points in the contribution workflow, from unreviewable PRs to fabricated security reports. Projects respond with three distinct mindsets — keep it out, let it in with conditions, or judge the contribution rather than the tool — put into practice through twelve concrete strategies: disclosure checkboxes, evidence requirements, PR limits, agent-facing instruction files, and more. Each choice carries real trade-offs, and copying the strictest policy you can find is usually a mistake. + +Whether you are a maintainer, PMC member, community builder, or OSPO practitioner, you will leave with a clear map of what other communities are doing, the trade-offs behind each option, and a practical starting point for the conversation in your own project. And before we close — one more thing: a first look at what actually happens when coding agents meet the rules you write. + +### Speakers: + + +
+ +Wenhao Yang: Peking University, PhD Candidate + +Wenhao Yang is a Ph.D. Candidate at the School of Computer Science, Peking University, and a member of Professor Minghui Zhou's Open Source Software Data Analytics Lab. His research focuses on open source software communities, software governance, and the impact of generative AI on collaborative development. He studies how open source projects adapt their contribution processes, policies, and review practices in response to emerging AI-assisted workflows. + + +
+ +Runzhi He: PhD Candidate, School of Computer Science, Peking University + +Runzhi He is currently a Ph.D. candidate at the School of Computer Science, Peking University, and a member of Professor Minghui Zhou's Open Source Software Data Analytics Lab. He received his B.S. degree from the School of Electronics Engineering and Computer Science (EECS), Peking University. His research mainly focuses on AI4SE and software supply chains. He can be contacted at rzhe@pku.edu.cn. \ No newline at end of file diff --git a/content/sessions/community-1210509.zh.md b/content/sessions/community-1210509.zh.md new file mode 100644 index 000000000..6c1989e08 --- /dev/null +++ b/content/sessions/community-1210509.zh.md @@ -0,0 +1,30 @@ +--- +title: "禁还是不禁？开源项目如何治理 GenAI 贡献" +date: "2026-08-07T14:30:00" +track: "community" +presenters: "杨文昊, 何润之" +stype: "中文演讲" +room: "玉泉山会议室" +--- + +生成式AI如今能编写代码、提交issue、发起拉取请求、评审变更，甚至在开源项目中报告漏洞。然而，生成成本变低了，并不代表审查成本也跟着降低。在日益沉重的维护负担下，许多项目已开始制定面向生成式AI的专门规则——写在 CONTRIBUTING 文件、安全策略、PR 模板和 AGENTS.md 中——从强制披露，一直到完全禁止。 + +为了弄清真正在发生什么，我们研读了 67 个最受关注的开源项目的 AI 规则，以及这些规则背后的事件与争论。整体图景远不止“禁还是不禁”那么简单。维护者们的忧虑集中在贡献流程中的七个压力点：从根本无法审查的 PR，到凭空捏造的安全报告。项目方则以三种截然不同的思路来应对——彻底拒之门外、有条件地放行，或只评判贡献本身而不问工具——并落实为十二种具体策略：披露复选框、证据要求、PR 提交限制、面向智能体的指令文件等等。每种选择都伴随着真实的权衡，照搬你所能找到的最严苛策略，通常是个错误。 + +无论你是维护者、PMC 成员、社区建设者还是 OSPO 从业者，你都将带着一张清晰的地图离开：了解其他社区在做什么、每种选择背后的权衡，以及在自己项目中启动相关讨论的实用起点。而在结束之前，还有一件事：首次揭示当编程智能体真的遭遇你写下的规则时，会发生什么。 + +### 讲师: + + +
+ +杨文昊：北京大学博士研究生 + +杨文昊是北京大学计算机学院的博士研究生，也是周明辉教授开源软件数据分析实验室的成员。他的研究方向涵盖开源软件社区、软件治理，以及生成式 AI 对协作开发的影响。他研究开源项目如何针对新兴的 AI 辅助工作流，调整其贡献流程、政策与评审实践。 + + +
+ +何润之：北京大学计算机学院博士研究生 + +何润之现为北京大学计算机学院博士研究生，也是周明辉教授开源软件数据分析实验室的成员。他本科毕业于北京大学信息科学技术学院（EECS），研究方向主要聚焦于 AI4SE 与软件供应链。联系邮箱：rzhe@pku.edu.cn。 \ No newline at end of file diff --git a/content/sessions/community-1211847.md b/content/sessions/community-1211847.md new file mode 100644 index 000000000..253ffaba0 --- /dev/null +++ b/content/sessions/community-1211847.md @@ -0,0 +1,36 @@ +--- +title: "Five Years into Open Source: Rethinking “Community Health” Through the Lens of a Storage Project" +date: "2026-08-08T14:30:00" +track: "community" +presenters: "Rui Su" +stype: "Chinese Session" +room: "Mtn YuQuan Hall" +--- + +Five years ago, we open-sourced a storage project. Like many teams, we believed at the time that if the code was good enough, a community would gradually form around it. Over the following five years, however, we came to realize that the growth of an open-source project is not only about code evolution. It is also about how a community is built, how it is perceived, and how it can remain sustainable over time. + +During these five years, we continuously observed and reflected on many signals commonly used to evaluate open-source projects: GitHub stars, issue activity, contributor numbers, user adoption, license choices, donations, and commercial support models. All of these signals are valuable, but they are also easy to oversimplify. For open-source infrastructure projects, community health does not always mean higher visibility or more activity. The real condition of a project often lies in more complex relationships: whether user adoption continues to grow, whether feedback is high quality, whether contributions are sustainable, whether the license builds trust, and whether the business model strengthens the community rather than weakening it. + +In this talk, I will share five years of practical experience from an open-source storage project, together with public data from sources such as OSS Insight, to discuss how open-source storage projects have changed in terms of community building and sustainability. The talk will focus on several questions: Which metrics should open-source projects really care about? Should infrastructure projects pay more attention to contributor growth or user adoption? How do different license choices and forms of commercial involvement affect community behavior and external perception? And in the reality of “open source + commercial operations,” how can a project build long-term trust? + +This is not a ranking of projects, nor a simple judgment of which model is better. It is a community retrospective based on real-world practice. I hope this talk can provide other open-source projects, especially infrastructure projects, with a more practical framework for observing and understanding community health. + +### Speakers: + + +
+ +Rui Su: Open-source advocate and co-founder of JuiceFS, a cloud-native distributed file system + +Rui Su is the co-founder of Juicedata and creator of JuiceFS, an open-source distributed file system built on object storage. He has spent the past seven years building storage infrastructure for AI, analytics, and cloud-native environments, and leading JuiceFS from a proprietary product into a globally adopted Apache 2.0 project. + +Rui speaks regularly on topics including AI/ML storage, Kubernetes-native data infrastructure, POSIX access over object storage, and open-source project strategy. He has presented at KubeCon 2023 Shanghai, SNIA SDC 2025, Community Over Code 2025, and KubeCon 2025 Atlanta. + +Selected talks: + +SNIA SDC 2025 — Why s3fs Fails in AI/ML and How to Achieve Scalable POSIX Access Anyway +https://www.youtube.com/watch?v=deQKBSAvt_M +Community Over Code 2025 — Why We Chose Openness: The Story Behind JuiceFS Switching to Apache 2.0 +https://www.youtube.com/watch?v=DtyJHVz4CJE +KubeCon 2023 Shanghai — How to Accelerate Model Training and Eliminate the I/O Bottleneck for the Cloud (in Chinese) +https://www.youtube.com/watch?v=pK9pUxRpBSM \ No newline at end of file diff --git a/content/sessions/community-1211847.zh.md b/content/sessions/community-1211847.zh.md new file mode 100644 index 000000000..a7817ff66 --- /dev/null +++ b/content/sessions/community-1211847.zh.md @@ -0,0 +1,36 @@ +--- +title: "开源五年：从一个存储项目重新审视'社区健康'" +date: "2026-08-08T14:30:00" +track: "community" +presenters: "Rui Su" +stype: "中文演讲" +room: "玉泉山会议室" +--- + +五年前，我们开源了一个存储项目。和许多团队一样，我们当时相信：只要代码足够好，社区就会自然而然地在它周围形成。然而在随后的五年里，我们逐渐意识到，一个开源项目的成长绝不仅仅是代码的演进，还关乎社区如何被建设、如何被认知，以及如何长期保持可持续。 + +在这五年里，我们持续观察并反思了许多常用于评估开源项目的信号：GitHub stars、issue 活跃度、贡献者数量、用户采用情况、许可证选择、捐赠，以及商业支持模式。所有这些信号都有价值，但也容易被过度简化。对于开源基础设施项目而言，社区健康并不总是意味着更高的可见度或更活跃的活动。一个项目的真实状况往往藏在更复杂的关系里：用户采用是否持续增长、反馈是否高质量、贡献是否可持续、许可证是否建立信任、商业模式是在强化社区还是在削弱它。 + +在本次演讲中，我将分享一个开源存储项目五年的实战经验，并结合来自 OSS Insight 等来源的公开数据，探讨开源存储项目在社区建设与可持续性方面发生了哪些变化。演讲将围绕几个问题展开：开源项目真正应该关注哪些指标？基础设施项目更应该看重贡献者增长，还是用户采用？不同的许可证选择与商业化介入形式，会如何影响社区行为与外部认知？而在"开源 + 商业化运营"的现实里，一个项目又该如何建立长期信任？ + +这既不是项目排行榜，也不是对哪种模式更好的简单评判。它是一次基于真实实践的社区复盘。我希望本次演讲能为其他开源项目（尤其是基础设施项目）提供一个更务实的框架，去观察和理解社区健康。 + +### 讲师: + + +
+ +Rui Su：开源倡导者，云原生分布式文件系统 JuiceFS 联合创始人 + +Rui Su 是 Juicedata 的联合创始人，也是 JuiceFS 的创建者——JuiceFS 是一个基于对象存储构建的开源分布式文件系统。过去七年里，他一直为 AI、分析和云原生环境构建存储基础设施，并带领 JuiceFS 从一款闭源产品走向全球采用的 Apache 2.0 项目。 + +Rui 经常围绕 AI/ML 存储、Kubernetes 原生数据基础设施、基于对象存储的 POSIX 访问，以及开源项目策略等主题发表演讲。他曾在 KubeCon 2023 上海、SNIA SDC 2025、Community Over Code 2025 以及 KubeCon 2025 亚特兰大发表演讲。 + +精选演讲： + +SNIA SDC 2025 —— 为什么 s3fs 在 AI/ML 场景下会失败，以及如何实现可扩展的 POSIX 访问 +https://www.youtube.com/watch?v=deQKBSAvt_M +Community Over Code 2025 —— 我们为何选择开放：JuiceFS 切换到 Apache 2.0 背后的故事 +https://www.youtube.com/watch?v=DtyJHVz4CJE +KubeCon 2023 上海 —— 如何为云端加速模型训练、消除 I/O 瓶颈（中文） +https://www.youtube.com/watch?v=pK9pUxRpBSM \ No newline at end of file diff --git a/content/sessions/community-1213694.md b/content/sessions/community-1213694.md new file mode 100644 index 000000000..ea294e8a4 --- /dev/null +++ b/content/sessions/community-1213694.md @@ -0,0 +1,31 @@ +--- +title: "Teaching the Apache Way with a little help from AI" +date: "2026-08-08T15:45:00" +track: "community" +presenters: "Justin Mclean" +stype: "English Session" +room: "Mtn YuQuan Hall" +--- + +Every ASF project leaves lessons behind. This talk shows how real Incubator case studies were turned into practical training using AI to identify patterns and summarise discussions, helping improve onboarding, mentoring, and community development. + + +Every ASF project has its own story. Some grow quickly, while others take years to find their rhythm, and still others fade away. Each leaves behind valuable lessons about community, governance, and collaboration. + +This talk shows how those lessons were turned into practical mentoring and training materials for the Apache Incubator. By analysing case studies from real projects and utilising AI tools to summarise patterns and draft content, we created new guides, quizzes, and interactive learning scenarios that help mentors and contributors learn from real-world experiences. + +We will share what went well, what did not, and how AI helped us quickly identify patterns and transform lengthy conversations into clear, useful learning materials. You will receive practical tips on how any open-source group can utilize AI to document its history, streamline onboarding, and facilitate more effective mentoring. + + +### Speakers: + + +
+ +Justin Mclean: ASF Director, VP ASF Incubator, Datastrato Community Manager + +Justin Mclean is a long-time contributor to the Apache Software Foundation (ASF) and currently serves as VP of the ASF Incubator and a member of the ASF Board of Directors. Over several years, he has contributed to evolving the Incubator’s governance, policies, and training programs, including documenting how projects operate in practice and analysing the Incubator’s history. Through this work, he has supported hundreds of open source projects in building sustainable and independent communities. + +He has extensive experience guiding projects from early incubation through to successful graduation, and has reviewed over a thousand open source releases. His work focuses on open source governance, community health, and the design of systems that enable collaboration at scale. + +Justin regularly speaks at international conferences on topics such as open source governance, licensing, and community building, sharing practical lessons drawn from real-world experience across a wide range of ASF projects and communities. \ No newline at end of file diff --git a/content/sessions/community-1213694.zh.md b/content/sessions/community-1213694.zh.md new file mode 100644 index 000000000..11cce53b9 --- /dev/null +++ b/content/sessions/community-1213694.zh.md @@ -0,0 +1,31 @@ +--- +title: "借助一点 AI 的帮助来传授 Apache 之道" +date: "2026-08-08T15:45:00" +track: "community" +presenters: "Justin Mclean" +stype: "英文演讲" +room: "玉泉山会议室" +--- + +每一个 ASF 项目都会留下经验教训。本次演讲展示如何把真实的 Incubator（孵化器）案例转化为实用的培训材料——借助 AI 来识别模式、总结讨论，从而帮助改进新人上手、导师辅导和社区发展。 + + +每一个 ASF 项目都有自己的故事。有的迅速成长，有的需要数年才找到节奏，还有的则逐渐淡出。每一个都留下了关于社区、治理与协作的宝贵经验。 + +本次演讲展示如何把这些经验教训转化为面向 Apache 孵化器的实用辅导与培训材料。通过分析真实项目的案例研究，并借助 AI 工具总结模式、起草内容，我们创建了新的指南、测验和互动式学习场景，帮助导师和贡献者从真实经历中学习。 + +我们将分享哪些做法奏效、哪些没有，以及 AI 如何帮助我们快速识别模式，并把冗长的对话转化为清晰、有用的学习材料。你将获得实用的建议，了解任何开源群体都可以如何利用 AI 来记录自己的历史、简化新人上手流程，并促进更高效的导师辅导。 + + +### 讲师: + + +
+ +Justin Mclean：ASF 董事，VP ASF Incubator，Datastrato 社区经理 + +Justin Mclean 是 Apache 软件基金会（ASF）的长期贡献者，目前担任 ASF 孵化器的 VP，同时也是 ASF 董事会成员。多年来，他参与了孵化器治理、政策和培训项目的演进，包括记录项目在实践中的真实运作方式，以及分析孵化器的历史。通过这些工作，他已支持数以百计的开源项目构建可持续、独立的社区。 + +他在引导项目从早期孵化到成功毕业方面经验丰富，并已审阅过上千次开源发布。他的工作聚焦于开源治理、社区健康，以及支撑大规模协作的系统设计。 + +Justin 经常在国际会议上围绕开源治理、许可证和社区建设等主题发表演讲，分享从众多 ASF 项目和社区的真实经验中提炼出的实践教训。 \ No newline at end of file diff --git a/content/sessions/community-1241720.md b/content/sessions/community-1241720.md new file mode 100644 index 000000000..5c55bafc2 --- /dev/null +++ b/content/sessions/community-1241720.md @@ -0,0 +1,23 @@ +--- +title: "When AI Becomes the Downstream: Rethinking Open Source Attribution, Licensing, and Long-Term Sustain" +date: "2026-08-07T15:30:00" +track: "community" +presenters: "Ian Xu" +stype: "Chinese Session" +room: "Mtn YuQuan Hall" +--- + +AI-assisted development and software agents are changing how open-source software is reused. Code can now be referenced, rewritten, or propagated through AI workflows without showing up in package manifests, forks, or traditional dependency graphs. + +This creates a new challenge for open-source communities: how do we preserve attribution, provenance, maintainer visibility, and long-term sustainability without restricting software freedom or blocking AI usage? + +This panel will discuss how open-source infrastructure may need to evolve in the agentic era, covering machine-readable attribution, dependency visibility, maintainer metadata, voluntary pledges, public accountability, and AI-era licensing. Using OpenSeed and the draft OpenSeed Agentic Open License 0.1 as a case study, we will explore how communities can keep open-source ecosystems visible, accountable, and sustainable as AI changes the meaning of downstream reuse. + +### Speakers: + + +
+ +Ian Xu: Founder of OpenSeed, Board Member of KAIYUANSHE, Co-founder of OpenBuild + +Founder of OpenSeed, Board Member of KAIYUANSHE, Co-founder of OpenBuild \ No newline at end of file diff --git a/content/sessions/community-1241720.zh.md b/content/sessions/community-1241720.zh.md new file mode 100644 index 000000000..8b595bc4c --- /dev/null +++ b/content/sessions/community-1241720.zh.md @@ -0,0 +1,23 @@ +--- +title: "当 AI 成为下游：重新思考开源的署名、许可与长期可持续性" +date: "2026-08-07T15:30:00" +track: "community" +presenters: "Ian Xu" +stype: "中文演讲" +room: "玉泉山会议室" +--- + +AI 辅助开发和软件 agent 正在改变开源软件被复用的方式。如今，代码可以经由 AI 工作流被引用、改写或传播，而不出现在包清单（package manifest）、fork 或传统的依赖图当中。 + +这给开源社区带来了一个新挑战：在不限制软件自由、也不封禁 AI 使用的前提下，我们如何保留署名（attribution）、来源（provenance）、维护者可见性以及长期可持续性？ + +本小组讨论将探讨在 agentic 时代，开源基础设施可能需要如何演进，话题涵盖机器可读的署名、依赖可见性、维护者元数据、自愿承诺（voluntary pledges）、公共问责，以及 AI 时代的许可证。我们将以 OpenSeed 及其草案版 OpenSeed Agentic Open License 0.1 为案例，探讨随着 AI 改变"下游复用"的含义，社区如何让开源生态保持可见、可问责且可持续。 + +### 讲师: + + +
+ +Ian Xu: OpenSeed 发起人, 开源社理事, OpenBuild Co-founder + +OpenSeed 发起人, 开源社理事, OpenBuild Co-founder \ No newline at end of file diff --git a/content/sessions/community-1290507.md b/content/sessions/community-1290507.md new file mode 100644 index 000000000..7d019bd41 --- /dev/null +++ b/content/sessions/community-1290507.md @@ -0,0 +1,24 @@ +--- +title: "Automate Everything Except the Maintainer" +date: "2026-08-07T15:00:00" +track: "community" +presenters: "Jarek Potiuk" +stype: "English Session" +room: "Mtn YuQuan Hall" +--- + +Every AI-for-open-source pitch promises the same thing: maintainers get their time back. But time was never the bottleneck — if it were, funding and contributor drives would have fixed burnout years ago. The real problem is that when a maintainer leaves, everything they knew leaves too: why that workaround exists, which "cleanup" detonates in production, what failed in 2019. Bus factor one. This talk makes the case that AI's real value isn't speed, it's transfer — turning what a maintainer knows into a SKILL an agent can run and a newcomer can read. Drawing on the decade from Breeze to Apache Magpie, I'll show where to draw the line between what you automate and what you don't, and why getting that line wrong builds a faster treadmill instead of a survivable project. + + +### Speakers: + + +
+ +Jarek Potiuk: Independent Open-Source Contributor and Advisor + +Independent Open-Source Contributor and Advisor, Committer and PMC member of Apache Airflow, Member of the Apache Software Foundation + +Jarek is an Engineer with a broad experience in many subjects - Open-Source, Cloud, Mobile, Robotics, AI, Backend, Developer Experience, but he also had a lot of non-engineering experience - running a company, being CTO, organizing big, international community events, technical sales support, pr and marketing advisory but also looking at legal aspect of licensing and building open-source communities are all under his belt. + +With the experience in very small and very big companies and everything in-between, Jarek found his place in Open-Source world, where his internal individual-contributor drive can be used to the uttermost of the potential. \ No newline at end of file diff --git a/content/sessions/community-1290507.zh.md b/content/sessions/community-1290507.zh.md new file mode 100644 index 000000000..3f81ffc31 --- /dev/null +++ b/content/sessions/community-1290507.zh.md @@ -0,0 +1,24 @@ +--- +title: "自动化一切，除了维护者" +date: "2026-08-07T15:00:00" +track: "community" +presenters: "Jarek Potiuk" +stype: "英文演讲" +room: "玉泉山会议室" +--- + +每一个面向开源的 AI 方案都在承诺同一件事：让维护者找回时间。但时间从来都不是瓶颈——如果是的话，资金支持和贡献者招募早就在多年前解决了倦怠问题。真正的问题在于，当一位维护者离开时，他所知道的一切也随之消失：那个变通方案为何存在、哪次"清理"会在生产环境中引发爆炸、2019 年到底什么失败了。巴士因子为一。本次演讲要论证的是，AI 真正的价值不在于速度，而在于传承——把维护者所掌握的知识转化为一种 Agent 可以执行、新人可以阅读的技能。结合从 Breeze 到 Apache Magpie 的十年经验，我将展示如何在"该自动化"与"不该自动化"之间划清界限，以及为什么这条线划错了，只会造出一台跑得更快的跑步机，而不是一个能够存续的项目。 + + +### 讲师: + + +
+ +Jarek Potiuk：独立开源贡献者与顾问 + +独立开源贡献者与顾问，Apache Airflow Committer 与 PMC 成员，Apache 软件基金会成员 + +Jarek 是一位工程师，在众多领域拥有丰富经验——开源、云计算、移动端、机器人、人工智能、后端、开发者体验；同时他也有大量非工程领域的经验——经营公司、担任 CTO、组织大型国际社区活动、技术销售支持、公关与营销咨询，以及研究许可证的法律层面和构建开源社区，这些都尽在他的掌握之中。 + +凭借在从极小到极大以及介于两者之间的各类公司中积累的经验，Jarek 在开源世界中找到了自己的位置，在这里，他作为个人贡献者的内在驱动力能够被发挥到极致。 \ No newline at end of file diff --git a/content/sessions/community-1293961.md b/content/sessions/community-1293961.md new file mode 100644 index 000000000..03acc1eec --- /dev/null +++ b/content/sessions/community-1293961.md @@ -0,0 +1,23 @@ +--- +title: "Free Software Won The Battle, But Not the War" +date: "2026-08-08T15:00:00" +track: "community" +presenters: "Shyam Sridhar" +stype: "English Session" +room: "Mtn YuQuan Hall" +--- + +The Free Software Movement was a line in the sand. In the 80s and 90s, as the industry was rushing to lock software behind proprietary licenses and call that progress, a handful of people steered a cultural movement with the goal to prevent this at any cost. Decades later, the results are in: ostensibly total victory in infrastructure norms. Open source is now the default way software is built. Almost all students learn development through it, and no CIO at any institution today needs to justify the decision to build on open source, and it has become so deeply normalised. + +If we look closely at what we won, every one of the four freedoms is a freedom over a thing you can hold - Run the program, study and adapt it, share it, improve and release it. But, software itself never existed in isolation, and this is especially true with every passing year. You can run free software end to end, every line auditable, every license clean, and it changes almost nothing about your actual position, because the restricted protocols beneath them, the opaque hardware that it runs on, the service it talks to, the platform that authenticates you, the API that can change its terms overnight, are all places where these freedoms are have been routed around. In fact, the irony hiding in plain sight is that GitHub: the social layer built around Git, where most open source projects live, is itself a closed system, where many of these freedoms are meaningfully constrained, and where the rules can change overnight. Moreover, the actors arriving at this layer are no longer just companies optimizing a subscription, they are Governments who are moving to wrap the whole stack inside national priorities. The lesson they learnt is simple: You don't need to own the code if you own the layer the code runs on. If this kills the four freedoms, will the entire edifice collapse? Did the free software movement lose the long game? + +This talk argues that the four freedoms secured freedom, but they do not secure freedom over the systems that software increasingly depends on. You can run entirely free software while still relying on infrastructure whose rules can be changed without your consent. The missing freedom is therefore a freedom from dependence on any single actor controlling the infrastructure beneath it. Call it the fifth freedom: the guarantee that the systems we rely on remain neutral and legible to all, applying the same rules regardless of who you are. Achieving that requires more than open source licenses. It requires systems engineered for credible neutrality, where no participant can permanently capture the commons. Yet those guarantees ultimately depend on the communities that defend them. Neutral protocols need communities willing to resist their capture, and communities need neutral protocols that cannot quietly be bent to private interests. Together, they form the technical and social foundations of infrastructure that remains trustworthy over time. + +### Speakers: + + +
+ +Shyam Sridhar: Ethereum Foundation + +Shyam Sridhar leads the Academic Secretariat at the Ethereum Foundation, serving as the interface between the Ethereum research ecosystem and academia by connecting open research questions with scholars and supporting independent research. \ No newline at end of file diff --git a/content/sessions/community-1293961.zh.md b/content/sessions/community-1293961.zh.md new file mode 100644 index 000000000..84a1f7707 --- /dev/null +++ b/content/sessions/community-1293961.zh.md @@ -0,0 +1,23 @@ +--- +title: "自由软件赢得了战役，却未赢得整场战争" +date: "2026-08-08T15:00:00" +track: "community" +presenters: "Shyam Sridhar" +stype: "英文演讲" +room: "玉泉山会议室" +--- + +自由软件运动曾是一道划在沙地上的界线。在 80 和 90 年代，随着整个行业争相将软件锁进专有许可证的牢笼、并称之为进步，一小群人引领了一场文化运动，目标是无论如何都要阻止这一切。几十年过去，结果已经揭晓：在基础设施的规范层面，这是一场表面上的彻底胜利。开源如今已成为软件开发默认的方式。几乎所有学生都通过它学习开发，今天任何机构的 CIO 都无需再为选择基于开源来构建而辩护，它已被深度常态化。 + +如果仔细审视我们所赢得的东西，四大自由中的每一项，都是针对一个你能够握在手中的事物的自由--运行程序、学习并改造它、分享它、改进并发布它。然而，软件本身从未孤立存在，而且这一点随着时间推移愈发明显。你可以端到端地运行自由软件，每一行代码都可审计，每一个许可证都干净无瑕，但这几乎不会改变你的实际处境，因为在它们之下受限制的协议、运行其上的不透明硬件、它所通信的服务、对你进行身份认证的平台、能够一夜之间改变条款的 API，都是这些自由被绕开的地方。事实上，一个明摆着的讽刺是：GitHub--围绕 Git 构建、承载着大多数开源项目的社会化层--本身就是一个封闭系统，在这里许多此类自由都受到实质性的限制，规则也能在一夜之间改变。更何况，进入这一层的参与者已不再只是优化订阅服务的公司，而是各国政府，它们正着手将整个技术栈纳入国家优先事项之中。他们学到的教训很简单：如果你拥有代码运行所在的那一层，你就不必拥有代码本身。如果这扼杀了四大自由，整座大厦是否会轰然倒塌？自由软件运动是否已经输掉了这场持久战？ + +本次演讲认为，四大自由确保了自由，但并未确保对软件日益依赖的那些系统的自由。你可以完全运行自由软件，却仍依赖着一套其规则可在未经你同意下被更改的基础设施。因此，所缺失的自由，是一种摆脱对控制其底层基础设施的任何单一行为者依赖的自由。姑且称之为第五项自由：一个保证--我们所依赖的系统对所有人保持中立且清晰可读，无论你是谁都一视同仁地适用相同的规则。实现这一点所需的远不止开源许可证。它需要为可信中立性而设计的系统，在其中没有任何参与者能够永久性地独占公共资源。然而，这些保证最终取决于捍卫它们的社区。中立的协议需要有愿意抵制其被俘获的社区，而社区也需要不会悄无声息地被私利所扭曲的中立协议。二者共同构成了那些能够长久保持可信的基础设施的技术与社会基石。 + +### 讲师: + + +
+ +Shyam Sridhar：以太坊基金会 + +Shyam Sridhar 在以太坊基金会领导学术秘书处，作为以太坊研究生态与学术界之间的接口，将开放的研究问题与学者对接，并支持独立研究。 \ No newline at end of file diff --git a/content/sessions/dataai-1185667.md b/content/sessions/dataai-1185667.md new file mode 100644 index 000000000..81b79aa6e --- /dev/null +++ b/content/sessions/dataai-1185667.md @@ -0,0 +1,32 @@ +--- +title: "From Lakehouse to Multimodal Data Lake: Rethinking Data Infrastructure for AI" +date: "2026-08-07T14:00:00" +track: "dataai" +presenters: "Zheng Yubin, Lili Ma" +stype: "Chinese Session" +room: "JingMing Hall" +--- + +The rapid rise of generative AI and machine learning workloads is fundamentally changing how data platforms are designed. Traditional data lakes and lakehouse architectures — built around large-scale scans and structured analytics — are increasingly challenged by new requirements such as vector search, multimodal data processing, and feature engineering pipelines. + +In this talk, we explore how data infrastructure is evolving in the AI era. Modern lakehouse table formats like Apache Iceberg have made schema evolution a well-solved problem. But AI workflows introduce a new challenge — Data Evolution: efficiently backfilling embeddings, recomputing features, and adding multimodal encodings without full table rewrites. + +We will also discuss emerging approaches for multimodal data management, including new table formats like Lance — built on the Apache Arrow type system — designed for AI workloads that unify structured data, embeddings, images, audio, and video in a single system. + +Finally, we will look at real-world architectural patterns such as Netflix's Media Data Lake and explore how open ecosystems are enabling a new generation of AI-native data platforms. + +### Speakers: + + +
+ +Zheng Yubin: AWS, Senior Developer Advocate + +Zheng Yubin, with over 20 years of experience in the ICT industry and digital transformation. Currently, serving as a Senior Developer Evangelist at AWS, specializing in Cloud Native, Cloud Security, and Generative AI. As the first female technical evangelist at AWS China, engaged with the developer community. As an architect with 18 years of experience, have provided consulting and technical implementation of solutions, including data center construction and software-defined data centers, for the finance, education, manufacturing, and high-tech industry. Leveraging industry expertise, offering technical guidance to developers, seeking mutual success. + + +
+ +Lili Ma: AWS, Senior Data Specialist Solutions Architect + +Lili Ma is a Data Specialist Solutions Architect at Amazon Web Services with over a decade of experience in data infrastructure research and product innovation. She started with Hadoop and Hive during her academic years, then moved through IBM DB2, the MPP data warehouse Greenplum, the compute-storage decoupled Apache HAWQ, and cloud-native databases Amazon Aurora and ElastiCache. She is a PMC member of the Apache HAWQ project and an early member of the Greenplum team. She has published multiple academic papers at international conferences including SIGMOD, GCC, SKG, and PDCAT, and holds several international patents. She has presented at ApacheCon, DTCC, KCD, and Greenplum community events on topics ranging from distributed database scalability to cloud-native architectures. \ No newline at end of file diff --git a/content/sessions/dataai-1185667.zh.md b/content/sessions/dataai-1185667.zh.md new file mode 100644 index 000000000..dd7c1e9df --- /dev/null +++ b/content/sessions/dataai-1185667.zh.md @@ -0,0 +1,32 @@ +--- +title: "从湖仓到多模态数据湖：重新思考面向 AI 的数据基础设施" +date: "2026-08-07T14:00:00" +track: "dataai" +presenters: "Zheng Yubin, Lili Ma" +stype: "中文演讲" +room: "静明厅" +--- + +生成式 AI 与机器学习工作负载的快速兴起，正在从根本上改变数据平台的设计方式。传统数据湖与湖仓架构——围绕大规模扫描与结构化分析而构建——正日益受到向量检索、多模态数据处理以及特征工程流水线等新需求的挑战。 + +在本次演讲中，我们将探讨数据基础设施在 AI 时代如何演进。Apache Iceberg 等现代湖仓表格式已使模式演进（schema evolution）成为已被妥善解决的问题。但 AI 工作流带来了一个全新挑战——数据演进（Data Evolution）：如何高效地回填嵌入（embedding）、重新计算特征、添加多模态编码，而无需对整张表进行重写。 + +我们还将讨论多模态数据管理的新兴方法，包括 Lance 等建立在 Apache Arrow 类型系统之上的新表格式——它专为 AI 工作负载设计，能在单一系统中统一管理结构化数据、嵌入、图像、音频与视频。 + +最后，我们将考察 Netflix 媒体数据湖（Media Data Lake）等真实架构模式，并探讨开放生态如何催生新一代 AI 原生的数据平台。 + +### 讲师: + + +
+ +Zheng Yubin：AWS 高级开发者布道师 + +Zheng Yubin 在 ICT 行业与数字化转型领域拥有超过 20 年经验。目前在 AWS 担任高级开发者布道师，专注于云原生、云安全与生成式 AI。作为 AWS 中国首位女性技术布道师，积极参与开发者社区。作为拥有 18 年经验的架构师，曾为金融、教育、制造和高科技行业提供数据中心建设、软件定义数据中心等解决方案的咨询与技术落地。依托行业专长，为开发者提供技术指导，寻求共赢。 + + +
+ +Lili Ma：AWS 高级数据专家解决方案架构师 + +Lili Ma 是 Amazon Web Services 的数据专家解决方案架构师，在数据基础设施研究与产品创新方面拥有十余年经验。她在学生时代即接触 Hadoop 与 Hive，此后历经 IBM DB2、MPP 数据仓库 Greenplum、存算分离的 Apache HAWQ，以及云原生数据库 Amazon Aurora 与 ElastiCache。她是 Apache HAWQ 项目的 PMC 成员，也是 Greenplum 团队的早期成员。她曾在 SIGMOD、GCC、SKG、PDCAT 等国际会议上发表多篇学术论文，并拥有多项国际专利。她曾在 ApacheCon、DTCC、KCD 及 Greenplum 社区活动中发表演讲，主题涵盖从分布式数据库可扩展性到云原生架构等多个方向。 \ No newline at end of file diff --git a/content/sessions/dataai-1186507.md b/content/sessions/dataai-1186507.md new file mode 100644 index 000000000..214541e3b --- /dev/null +++ b/content/sessions/dataai-1186507.md @@ -0,0 +1,41 @@ +--- +title: "From MCP to Dynamic Ontology: How Enterprises Can Choose the Right ChatBI and Data Agent Architecture on Apache Doris" +date: "2026-08-07T15:00:00" +track: "dataai" +presenters: "奕嘉 苏" +stype: "Chinese Session" +room: "JingMing Hall" +--- + +When enterprises begin building ChatBI or Data Agents, the hardest question is often not whether a large language model can generate SQL. It is how to choose a cost-effective architecture that can evolve with the company's existing data foundation, business complexity, and risk profile. + +A team that has just completed a real-time data warehouse and wants to validate the value of conversational analytics should not use the same architecture as an enterprise that already has a mature metrics platform and needs to coordinate analysis across sales, supply chain, production, and service. +Direct NL2SQL is not inherently wrong. When the data scope is limited, business definitions are clear, and access is read-only and low-risk, it can be the fastest way to validate value. +The problem begins when enterprises take an architecture designed for a proof of concept and expose it directly to hundreds or thousands of tables in a complex production environment, leaving schema discovery, business semantics, permissions, security, cost, and result correctness to a single probabilistic model. + +This session will first explain why Apache Doris is well suited to serve as the trusted data and execution foundation for enterprise ChatBI and Data Agents. In addition to unified SQL execution for real-time analytics, Doris can use views and materialized views to deliver Agent-facing data products, establish execution boundaries through access control, masking, workload management, and auditing, and allow Agent platforms such as Dify to discover metadata, inspect SQL execution plans, and run read-only queries through the official Doris MCP Server. + +On that foundation, I will present a route-selection framework based on an enterprise's current capabilities: + +1. **Exploratory PoC: raw schemas + Doris MCP + a read-only Agent.** This route offers a low-cost, fast-feedback way to validate demand in a small, low-risk scope. It is not suitable for exposing thousands of raw tables directly to a model, and an executable SQL statement must not be mistaken for a correct business answer. +2. **Data-product route: domain-oriented wide tables, views, or materialized views + a constrained query surface.** This route suits enterprises with a basic data warehouse that want to improve the accuracy of high-frequency operational analysis. Data modeling reduces the model's search space and is often more effective than adding more prompt instructions, but it also introduces data-product development and maintenance costs. +3. **Metrics-layer reuse: an existing metrics platform or semantic API + LLM intent mapping.** This route suits enterprises that already have well-governed metric definitions, permissions, and service interfaces. The model primarily understands the question and maps parameters instead of reinventing metric SQL. It provides stable definitions but is better suited to standardized metric Q&A than to long-tail analysis involving cross-domain objects and states. +4. **Dynamic-ontology-oriented semantic layer: unified modeling of business objects, relationships, events, states, metrics, and permissions.** This route fits complex cross-domain analysis, evolving semantics, and scenarios that require explanation, regression testing, and auditing. It is not the starting point for every enterprise. Rather, it becomes valuable when data products and metric governance have reached sufficient maturity and a more advanced semantic model and execution-control layer is required for complex Data Agents. + +To test these judgments, we used the same Apache Doris data snapshot, model, and official MCP Server with an enterprise data environment containing more than 1,000 tables. We evaluated the approaches with 100 executable questions and 20 additional questions covering clarification, security, and capability boundaries. + +Finally, I will use YuanYi's dynamic-ontology-oriented semantic layer as a concrete implementation example for complex enterprise scenarios: + +`Natural-language question → Semantic Catalog / Candidate Graph → YuanYi Logic Plan (YLP) → Query Guard → Doris SQL → Answer Contract → Trace / Audit` + +Dynamic ontology is neither another layer floating above the semantic layer nor simply a longer prompt. It is a higher-order modeling kernel inside a stronger semantic layer. It extends metric semantics to business objects, relationships, events, states, and controlled actions. Through intermediate representations, compilation, permissions, and regression mechanisms, the large language model is responsible for understanding and planning, while system boundaries remain responsible for facts, security, and traceability. + +Attendees will leave with a route-selection map that can be used directly in enterprise architecture discussions, a phased implementation playbook from PoC to production, and a practical understanding of the costs, benefits, limitations, and common pitfalls of each route. The session will not argue that every enterprise should build a heavyweight platform from day one. Instead, it will help teams choose the smallest viable architecture for their current stage while preserving the right extension points for what comes next. + +### Speaker: + +
+ +Yijia Su: Founder of YuanYi Intelligence, Apache Doris Committer, Apache Doris 2025 MVP, Doris-MCP Contributor, and PowerData Community Initiator + +Yijia has long contributed to the Apache Doris community and to enterprise implementations of real-time data warehouses, ChatBI, and Data Agents. His current work focuses on enterprise semantic layers with dynamic ontology as the modeling kernel, and on explainable, regression-tested, and auditable infrastructure for Data Agents. \ No newline at end of file diff --git a/content/sessions/dataai-1186507.zh.md b/content/sessions/dataai-1186507.zh.md new file mode 100644 index 000000000..16d4f6eee --- /dev/null +++ b/content/sessions/dataai-1186507.zh.md @@ -0,0 +1,41 @@ +--- +title: "从 MCP 到动态本体：企业如何基于 Apache Doris 选择 ChatBI 与 Data Agent 技术路线" +date: "2026-08-07T15:00:00" +track: "dataai" +presenters: "奕嘉 苏" +stype: "中文演讲" +room: "静明厅" +--- + +当企业开始建设 ChatBI 或 Data Agent，真正困难的往往不是“大模型会不会生成 SQL”，而是如何根据自己的数据基础、业务复杂度和风险等级，选择一条成本可控、能够逐步演进的技术路线。 + +一个刚完成实时数仓建设、只想验证自然语言问数价值的团队，和一个已经拥有成熟指标平台、需要跨销售、供应链、生产与服务协同的企业，显然不应该采用同一套架构。 +直接 NL2SQL 并非天然错误：在数据范围有限、口径清晰、只读低风险的场景中，它反而是验证价值最快的方式。 +问题在于，很多企业把适合 PoC 的方案直接推向数百、上千张表和复杂生产环境，于是将 Schema 发现、业务语义、权限、安全、成本和结果正确性同时交给了一个概率模型。 + +本次演讲将首先回答：为什么 Apache Doris 适合作为企业 ChatBI 与 Data Agent 的事实和执行底座。Doris 不仅提供面向实时分析的统一 SQL 执行能力，还可以通过视图与物化视图沉淀面向 Agent 的数据产品，通过权限、脱敏、负载管理和审计建立执行边界，并借助官方 Doris MCP Server 让 Dify 等 Agent 平台以通用工具方式完成元数据发现、SQL Explain 与只读查询。 + +在此基础上，我将给出一套面向企业现状的路线选择框架： + +1. **探索型 PoC：原始 Schema + Doris MCP + 只读 Agent。** 适合小范围、低风险地验证需求，成本低、反馈快；但不适合把上千张原始表直接暴露给模型，更不能把“SQL 执行成功”误认为“业务答案正确”。 +2. **数据产品路线：主题宽表、视图或物化视图 + 查询范围约束。** 适合已经有基本数仓能力、希望先提升高频经营分析准确率的企业。它通过数据建模缩小模型搜索空间，通常比继续堆 Prompt 更有效，但会带来数据产品建设和维护成本。 +3. **指标层复用路线：既有指标平台或语义 API + LLM 意图映射。** 适合已经拥有较完善指标口径、权限和服务接口的企业。模型主要负责理解问题和映射参数，而不是重新发明指标 SQL；优势是口径稳定，局限是更适合标准指标问答，对跨域对象关系、状态和长尾分析的覆盖有限。 +4. **动态本体导向的语义层：业务对象、关系、事件、状态、指标与权限的统一建模。** 适合跨域分析复杂、语义持续变化，并要求解释、回归和审计的场景。它不是所有企业的起点，而是在数据产品和指标治理达到一定基础后，为复杂 Data Agent 提供更高阶的语义建模与执行控制。 + +为了验证这些判断，我们在同一 Apache Doris 数据快照、同一模型和同一官方 MCP Server 上，借助一份包含 1000 余张表的企业数据环境，并使用 100 道可执行题和 20 道澄清、安全及能力边界题进行对照测试。 + +最后，我将以元一动态本体化语义层的实践为例，介绍一条更适合复杂企业场景的实现链路： + +`自然语言问题 → Semantic Catalog / Candidate Graph → YuanYi Logic Plan（YLP）→ Query Guard → Doris SQL → Answer Contract → Trace / Audit` + +动态本体不是悬浮在语义层之上的另一层，也不是一个更长的 Prompt。它是更强语义层内部的高阶建模内核：把指标语义扩展到业务对象、关系、事件、状态和受控动作，并通过中间表示、编译、权限与回归机制，让大模型负责理解和规划，让系统边界负责事实、安全与可追溯性。 + +大家将带走一张可直接用于企业方案讨论的路线选择图、一套从 PoC 到生产的分阶段实施建议，以及对不同路线成本、收益、天花板和常见误区的真实判断。演讲不会主张所有企业都建设一套重型平台，而是帮助团队先选择当前最小可行架构，再为下一阶段留下正确的演进接口。 + +### 讲师: + +
+ +苏奕嘉：元一智能创始人、Apache Doris Committer、Apache Doris 2025 MVP、Doris-MCP 贡献者、PowerData 社区发起人 + +长期参与 Apache Doris 社区建设与企业实时数仓、ChatBI、Data Agent 落地。目前专注于以动态本体为建模内核的企业语义层，以及可解释、可回归、可审计的 Data Agent 基础设施。 \ No newline at end of file diff --git a/content/sessions/dataai-1206062.md b/content/sessions/dataai-1206062.md new file mode 100644 index 000000000..d4902d85a --- /dev/null +++ b/content/sessions/dataai-1206062.md @@ -0,0 +1,37 @@ +--- +title: "Vector Lakebase: Rethinking Data Infrastructure for AI-Native Applications" +date: "2026-08-08T16:15:00" +track: "dataai" +presenters: "Li Liu" +stype: "Chinese Session" +room: "JingMing Hall" +--- + +Description +As AI applications move from prototypes to production, traditional data infrastructure is starting to show its limits. Systems designed for analytics or transactions struggle to support modern AI workloads — especially vector search, hybrid filtering, and multi-modal data at scale. + +In this talk, we present Milvus and the concept of a Vector Lakebase — an AI-native data platform that unifies vector search with lakehouse storage. Instead of building yet another standalone vector database, Vector Lakebase brings indexing and retrieval directly to data lakes such as Iceberg and Paimon. + +We’ll dive into the architecture and key design decisions behind Milvus, including: + +How to enable efficient vector search directly on Iceberg/Paimon tables without data duplication +Zero-copy data access: separating compute and storage while avoiding costly data movement +Snapshot-aligned indexing: keeping vector indexes consistent with lakehouse metadata +Handling production challenges like multi-tenancy, filtering, and cold-start performance + +We’ll also share real-world patterns where this architecture matters most: + +Agentic search systems that require consistent, high-recall retrieval +Data discovery across large, heterogeneous datasets +Data deduplication and similarity-based data management at scale + +Finally, we’ll discuss how AI-native data infrastructure is evolving — and why integrating vector search into the lake, rather than building isolated systems, is becoming the dominant pattern. + +### Speakers: + + +
+ +Li Liu: Director of Engineering + +Li Liu is Director of Engineering at Zilliz and Technical Lead for Milvus. With many years of experience in databases and big-data processing, he now leads the development and maintenance of Zilliz’s core database systems. Previously, he was a Senior Engineer at Meta, where he designed and built the company’s streaming data framework for advertising. Li holds a Master of Information Technology degree from Carnegie Mellon University. \ No newline at end of file diff --git a/content/sessions/dataai-1206062.zh.md b/content/sessions/dataai-1206062.zh.md new file mode 100644 index 000000000..be8150503 --- /dev/null +++ b/content/sessions/dataai-1206062.zh.md @@ -0,0 +1,37 @@ +--- +title: "向量湖仓（Vector Lakebase）：重新思考面向 AI 原生应用的数据基础设施" +date: "2026-08-08T16:15:00" +track: "dataai" +presenters: "刘力" +stype: "中文演讲" +room: "静明厅" +--- + +描述 +随着 AI 应用从原型走向生产，传统数据基础设施的局限性开始显现。为分析或事务设计的系统，难以支撑现代 AI 工作负载——尤其是大规模的向量检索、混合过滤与多模态数据。 + +在本次演讲中，我们将介绍 Milvus 以及"向量湖仓"（Vector Lakebase）的概念——这是一个 AI 原生的数据平台，将向量检索与湖仓存储统一起来。Vector Lakebase 不再是构建又一个独立的向量数据库，而是把索引与检索能力直接带到 Iceberg、Paimon 等数据湖之上。 + +我们将深入 Milvus 背后的架构与关键设计决策，包括： + +如何在不复制数据的前提下，直接在 Iceberg/Paimon 表上实现高效向量检索 +零拷贝数据访问：分离计算与存储，同时避免代价高昂的数据搬运 +快照对齐的索引：让向量索引与湖仓元数据保持一致 +应对多租户、过滤、冷启动性能等生产级挑战 + +我们还会分享该架构最能发挥价值的一些真实场景： + +需要一致、高召回率检索的 Agentic 检索系统 +跨越大型异构数据集的数据发现 +大规模场景下的数据去重与基于相似度的数据管理 + +最后，我们将讨论 AI 原生数据基础设施的演进方向——以及为何"将向量检索融入数据湖"而非构建孤立系统，正在成为主导范式。 + +### 讲师: + + +
+ +刘力：工程总监 + +刘力现任 Zilliz 工程总监兼 Milvus 技术负责人。他在数据库和大数据处理领域拥有多年经验，目前负责 Zilliz 核心数据库系统的开发与维护。此前，他曾任 Meta 高级工程师，设计并构建了该公司面向广告业务的流式数据框架。刘力拥有卡内基梅隆大学信息技术硕士学位。 \ No newline at end of file diff --git a/content/sessions/dataai-1207761.md b/content/sessions/dataai-1207761.md new file mode 100644 index 000000000..e6c3a22e3 --- /dev/null +++ b/content/sessions/dataai-1207761.md @@ -0,0 +1,21 @@ +--- +title: "Building a Palantir-like Data & AI Platform with the Apache Stack" +date: "2026-08-07T16:15:00" +track: "dataai" +presenters: "Zhangjian He" +stype: "Chinese Session" +room: "JingMing Hall" +--- + +Modern organizations are increasingly seeking a unified platform that combines data integration, governance, lineage, and operational applications—similar to what platforms like Palantir Foundry provide. However, building such a system from scratch often appears prohibitively complex. + +This session presents a pragmatic, engineering-driven approach to constructing a “Foundry-like” platform using the Apache ecosystem. + +### Speakers: + + +
+ +Zhangjian He: Huawei Cloud Senior Engineer | Huawei Cloud Open Source Team Member + +Open-source enthusiast and experienced engineer. Since starting software development in 2017, I have played a key role in leading the cloud-native transformation of Huawei Cloud's IoT platform and the development of an industrial IoT platform. I serve as an Apache BookKeeper/ServiceComb PMC member, Apache Pulsar Committer, and CNCF openGemini Maintainer. I specialize in coding, software engineering, open source, and ecosystem development, driving best practices and development standards. \ No newline at end of file diff --git a/content/sessions/dataai-1207761.zh.md b/content/sessions/dataai-1207761.zh.md new file mode 100644 index 000000000..be316313e --- /dev/null +++ b/content/sessions/dataai-1207761.zh.md @@ -0,0 +1,21 @@ +--- +title: "用 Apache 技术栈打造类 Palantir 的数据与 AI 平台" +date: "2026-08-07T16:15:00" +track: "dataai" +presenters: "Zhangjian He" +stype: "中文演讲" +room: "静明厅" +--- + +越来越多的现代企业希望拥有一个集数据集成、治理、血缘与业务应用于一体的统一平台——类似于 Palantir Foundry 这类平台所提供的能力。然而，从零开始构建这样一套系统，往往显得复杂得令人望而却步。 + +本次演讲将介绍一种务实、以工程为驱动的方法，讲解如何基于 Apache 生态构建一个"类 Foundry"的平台。 + +### 讲师: + + +
+ +Zhangjian He：华为云高级工程师 | 华为云开源团队成员 + +开源爱好者，经验丰富的工程师。自 2017 年开始软件开发以来，我在华为云物联网平台的云原生转型以及工业物联网平台的开发中扮演了关键角色。我担任 Apache BookKeeper/ServiceComb PMC 成员、Apache Pulsar Committer，以及 CNCF openGemini Maintainer。我专长涵盖编码、软件工程、开源与生态建设，致力于推动最佳实践与开发标准。 \ No newline at end of file diff --git a/content/sessions/dataai-1210242.md b/content/sessions/dataai-1210242.md new file mode 100644 index 000000000..804aa87b4 --- /dev/null +++ b/content/sessions/dataai-1210242.md @@ -0,0 +1,21 @@ +--- +title: "GeaFlow｜How Streaming Graph Computing Reshapes the Memory Operating System of AI Agents" +date: "2026-08-08T14:00:00" +track: "dataai" +presenters: "Litao Lin" +stype: "Chinese Session" +room: "JingMing Hall" +--- + +By 2026, the memory architecture of AI Agents is undergoing a paradigm shift from "vector retrieval" to "graph reasoning." Memory frameworks have validated an industry consensus: pure vector databases fail to capture the evolutionary relationships between entities and multi-hop causal chains, while dumping complex data wholesale into large language models is both expensive and unsustainable—this is precisely the critical bottleneck preventing Agents from evolving from "having memory" to "being capable of reasoning." Meanwhile, over 95% of enterprise-grade Agent deployments are moving toward private cloud or on-premise environments. The business intelligence and customer data distilled within Agent memory makes data sovereignty a non-negotiable red line, urgently requiring a memory infrastructure that can be autonomously controlled. +This presentation proposes a new perspective: streaming graph computing engines are inherently the operating system kernel for Agent memory. Apache GeaFlow (Incubating)'s incremental computation capabilities correspond to the real-time extraction and merging of memories—where each new conversation triggers only local subgraph updates; dynamic graph state management corresponds to temporal memory graphs—where vertices store semantic embeddings, edges store relationships and temporal metadata, and properties store key textual descriptions; distributed snapshot mechanisms correspond to memory consistency guarantees in enterprise-scale multi-Agent collaboration scenarios. Through the MCP (Model Context Protocol) standard, GeaFlow serves as a plug-and-play memory backend for Agent frameworks. +In technical practice, we will demonstrate how to build an enterprise knowledge assistant with "long-term memory" based on GeaFlow, showcasing the dynamic construction and reasoning processes of memory graphs, and outline the evolution roadmap for GraphMemory—encompassing streaming graph neural network inference, proactive memory synthesis, and natural forgetting mechanisms based on graph algorithms. + +### Speakers: + + +
+ +Litao Lin: Apache GeaFlow (Incubating) Committer + +Apache GeaFlow (Incubating) Committer. As a core member of the project, he participated in the architecture design and development of the GeaFlow graph computing engine from scratch, with a particular focus on the design and implementation of the graph computing DSL and the evolution of data intelligence technologies. He is deeply involved in open source community activities and currently works at Ant Group. \ No newline at end of file diff --git a/content/sessions/dataai-1210242.zh.md b/content/sessions/dataai-1210242.zh.md new file mode 100644 index 000000000..e1a51b9c7 --- /dev/null +++ b/content/sessions/dataai-1210242.zh.md @@ -0,0 +1,21 @@ +--- +title: "GeaFlow｜流式图计算如何重塑 AI Agent 的记忆操作系统" +date: "2026-08-08T14:00:00" +track: "dataai" +presenters: "Litao Lin" +stype: "中文演讲" +room: "静明厅" +--- + +到 2026 年，AI Agent 的记忆架构正在经历从"向量检索"到"图推理"的范式转换。记忆框架已经印证了一个行业共识：纯向量数据库无法刻画实体间的演化关系与多跳因果链，而把复杂数据一股脑倒给大语言模型既昂贵又不可持续——这恰恰是阻碍 Agent 从"拥有记忆"迈向"具备推理能力"的关键瓶颈。与此同时，超过 95% 的企业级 Agent 部署正走向私有云或本地化环境。Agent 记忆中所沉淀的业务情报与客户数据，使数据主权成为不可逾越的红线，亟需一套可自主掌控的记忆基础设施。 +本次演讲提出一个新视角：流式图计算引擎天然就是 Agent 记忆的操作系统内核。Apache GeaFlow（孵化中）的增量计算能力，对应记忆的实时抽取与合并——每一次新对话只触发局部子图的更新；动态图状态管理对应时序记忆图——顶点存储语义嵌入，边存储关系与时序元数据，属性存储关键文本描述；分布式快照机制则对应企业级多 Agent 协作场景下的记忆一致性保障。借助 MCP（模型上下文协议）标准，GeaFlow 可作为 Agent 框架即插即用的记忆后端。 +在技术实践中，我们将演示如何基于 GeaFlow 构建具备"长期记忆"的企业知识助手，展示记忆图的动态构建与推理过程，并勾勒 GraphMemory 的演进路线图——涵盖流式图神经网络推理、主动式记忆合成，以及基于图算法的自然遗忘机制。 + +### 讲师: + + +
+ +Litao Lin：Apache GeaFlow（孵化中）Committer + +Apache GeaFlow（孵化中）Committer。作为项目核心成员，他从零参与 GeaFlow 图计算引擎的架构设计与开发，尤其专注于图计算 DSL 的设计与实现以及数据智能技术的演进。他深度参与开源社区活动，目前就职于蚂蚁集团。 \ No newline at end of file diff --git a/content/sessions/dataai-1210452.md b/content/sessions/dataai-1210452.md new file mode 100644 index 000000000..e724c71ce --- /dev/null +++ b/content/sessions/dataai-1210452.md @@ -0,0 +1,25 @@ +--- +title: "Xiaomi's HDFS Cloud Migration Practice: Building a Unified Storage for Data + AI" +date: "2026-08-08T15:00:00" +track: "dataai" +presenters: "朝晖 王" +stype: "Chinese Session" +room: "JingMing Hall" +--- + +With the rapid development of Xiaomi's business needs, particularly the ongoing implementation in scenarios like Generative AI and autonomous driving, the underlying technology stack is continuously evolving towards big data and cloud-native directions. This evolution places multiple demands on the storage system, including large capacity, high performance, low cost, and multi-protocol access. Simultaneously, business data often needs to flow across multiple systems such as object storage, high-performance file storage, HDFS, and local disks, leading to issues like complex pipelines, elevated migration and operational costs, and uneven resource utilization. To address these challenges, we have gradually implemented a unified file storage solution in our production environment to support data access, management, and distribution across various scenarios. This practice currently covers 40 clusters, supporting the management of hundreds of billions of files and a storage scale of hundreds of petabytes. +This presentation will focus on this solution, detailing the following key aspects: +Unified Architecture Design:​ The overall division of labor among the client, control plane, and data plane, along with a tiered design comprising the capacity layer, cache layer, and performance layer; client-side support for various access methods including POSIX, HDFS, S3, CSI Driver, and Python SDK. +Storage Practice for the Full AI Pipeline:​ How to meet the requirements of different stages—raw data, training data, and model file distribution—and optimize read-intensive, write-rare scenarios through caching mechanisms. +Cloud Migration Practice for the Apache Big Data Ecosystem:​ How to achieve compute-storage separation for cloud migration based on Apache Hadoop/HDFS compatibility, support the conversion of Apache Hive tables to Apache Iceberg, and enable data lakehouse scenarios with technologies like Apache Paimon. +Hot Data Caching and Cloud-Native Practice:​ How to improve computational efficiency through automatic caching of hot tables, achieve compute-storage separation for cloud-migrated services via CSI mounting, and reduce disk redundancy and accelerate model loading through shared data and cache distribution. +Through these practices, we aim to share insights on implementing unified storage in Data + AI environments and how to balance migration complexity, system scalability, access performance, and overall cost. + +### Speakers: + + +
+ +朝晖 王: Xiaomi Software R&D Engineer + +HDFS/Ozone/JuiceFS contributor \ No newline at end of file diff --git a/content/sessions/dataai-1210452.zh.md b/content/sessions/dataai-1210452.zh.md new file mode 100644 index 000000000..244753b0d --- /dev/null +++ b/content/sessions/dataai-1210452.zh.md @@ -0,0 +1,25 @@ +--- +title: "小米 HDFS 云迁移实践：打造面向 Data + AI 的统一存储" +date: "2026-08-08T15:00:00" +track: "dataai" +presenters: "朝晖 王" +stype: "中文演讲" +room: "静明厅" +--- + +随着小米业务需求的快速发展，尤其是生成式 AI、自动驾驶等场景的持续推进，底层技术栈不断向大数据和云原生方向演进。这种演进对存储系统提出了多重诉求：大容量、高性能、低成本，以及多协议访问。与此同时，业务数据常常需要在对象存储、高性能文件存储、HDFS 和本地磁盘等多个系统之间流转，由此带来流水线复杂、迁移与运维成本高企、资源利用不均等问题。为应对这些挑战，我们逐步在生产环境中落地了一套统一文件存储方案，以支撑各类场景下的数据访问、管理与分发。该实践目前覆盖 40 个集群，支撑千亿级文件的管理与数百 PB 级的存储规模。 +本次演讲将围绕这一方案展开，详细讲解以下关键方面： +统一架构设计：客户端、控制面与数据面之间的整体分工，以及由容量层、缓存层和性能层构成的分层设计；客户端支持 POSIX、HDFS、S3、CSI Driver 和 Python SDK 等多种访问方式。 +面向 AI 全流水线的存储实践：如何满足原始数据、训练数据和模型文件分发等不同阶段的需求，并通过缓存机制优化"读多写少"的场景。 +Apache 大数据生态的云迁移实践：如何基于 Apache Hadoop/HDFS 兼容性实现云迁移所需的存算分离，支持 Apache Hive 表向 Apache Iceberg 转换，并借助 Apache Paimon 等技术实现数据湖仓场景。 +热数据缓存与云原生实践：如何通过热点表的自动缓存提升计算效率，通过 CSI 挂载为云迁移业务实现存算分离，并通过共享数据与缓存分发降低磁盘冗余、加速模型加载。 +通过这些实践，我们希望分享在 Data + AI 环境下落地统一存储的洞察，以及如何在迁移复杂度、系统可扩展性、访问性能与总体成本之间取得平衡。 + +### 讲师: + + +
+ +朝晖 王：小米软件研发工程师 + +HDFS/Ozone/JuiceFS 贡献者 \ No newline at end of file diff --git a/content/sessions/dataai-1210506.md b/content/sessions/dataai-1210506.md new file mode 100644 index 000000000..c45e81fca --- /dev/null +++ b/content/sessions/dataai-1210506.md @@ -0,0 +1,25 @@ +--- +title: "Real-Time Sports Commentary as a Lens into Flink's New Paradigm for Streaming AI" +date: "2026-08-08T14:30:00" +track: "dataai" +presenters: "Xintong Song" +stype: "Chinese Session" +room: "JingMing Hall" +--- + +Today's AI isn't slow — a single request returns results in seconds. But this "online real-time" is fundamentally passive inference triggered by requests, while the real world produces data as a continuous stream of events. When AI needs to perform continuous, incremental, and stateful processing over real-time data streams, we enter a different paradigm — Streaming AI. It raises distinctive challenges: treating multimodal data as a first-class citizen in streaming pipelines; supporting locally-executed computation such as video decoding, image processing, and small-model inference with GPUs inside the streaming runtime, rather than detouring through remote services; integrating agent-style multi-step reasoning, tool use, and context and memory management with event-driven stream processing; and leveraging stream processing's heritage in joins, aggregations, state management, and temporal guarantees, rather than rebuilding it from scratch. + +Flink is addressing these challenges through a series of advancements: multimodal data processing for audio, video, images, and vectors as first-class data; GPU scheduling for local computation within the Flink runtime; Flink Agents that natively embed agents with context management, memory, and tool invocation into stream processing; and a DataFrame API that offers a unified, concise interface for AI and data engineers. Combined with Flink's core strengths in streaming state and temporal semantics, these advancements define Flink's new role — the unified foundation for Streaming AI. + +This talk uses a real-time AI sports commentary system built on Flink as an end-to-end case study: from live-stream frame extraction and visual understanding, to context-aware commentary generation, to TTS synthesis synchronized with the broadcast. A seemingly simple "AI commentator" exercises every one of these capabilities at once. + +Attendees will leave with a clear view of Streaming AI's core challenges, Flink's latest AI capabilities, a reusable multimodal pipeline architecture, and engineering insights for balancing latency, cost, and model quality in production. + +### Speakers: + + +
+ +Xintong Song: Apache Flink PMC Member, Staff Software Engineer @ Alibaba Cloud + +Xintong is an Apache Flink PMC member and a main promoter of the Flink Agents sub-project. He is also a Staff Software Engineer at Alibaba Cloud, leading a team that works on Flink's AI and Lakehouse integrations. Prior to that, he received a Ph.D. degree in computer science from Peking University. \ No newline at end of file diff --git a/content/sessions/dataai-1210506.zh.md b/content/sessions/dataai-1210506.zh.md new file mode 100644 index 000000000..73c0e453a --- /dev/null +++ b/content/sessions/dataai-1210506.zh.md @@ -0,0 +1,25 @@ +--- +title: "从实时体育解说看 Flink 流式 AI 新范式" +date: "2026-08-08T14:30:00" +track: "dataai" +presenters: "宋辛童" +stype: "中文演讲" +room: "静明厅" +--- + +今天的 AI 并不慢——单个请求几秒内就能返回结果。但这种"在线实时"本质上是由请求触发的被动推理，而真实世界产生的数据却是一连串持续的事件流。当 AI 需要在实时数据流上做连续、增量且有状态的处理时，我们便进入了一种全新范式——流式 AI（Streaming AI）。它带来一系列独特挑战：把多模态数据作为流处理流水线中的一等公民；在流处理运行时内支持本地执行的计算，例如视频解码、图像处理以及在运行时中借助 GPU 进行小模型推理，而非绕道远程服务；将 Agent 式的多步推理、工具调用、上下文与记忆管理与事件驱动的流处理相融合；并复用流处理在 join、聚合、状态管理和时间语义方面的深厚积累，而非从零重建。 + +Flink 正通过一系列进展来应对这些挑战：把音频、视频、图像和向量等多模态数据处理为一等数据；为 Flink 运行时内的本地计算提供 GPU 调度；Flink Agents 将带有上下文管理、记忆和工具调用的 Agent 原生嵌入流处理之中；以及为 AI 与数据工程师提供统一、简洁接口的 DataFrame API。这些进展与 Flink 在流式状态和时间语义上的核心优势相结合，共同定义了 Flink 的新角色——流式 AI 的统一底座。 + +本次演讲将以一个基于 Flink 构建的实时 AI 体育解说系统作为端到端案例：从直播帧抽取与视觉理解，到结合上下文的解说生成，再到与播出画面同步的 TTS 合成。一个看似简单的"AI 解说员"，会同时用到上述每一项能力。 + +听众将带走对流式 AI 核心挑战的清晰认识、Flink 最新的 AI 能力、一套可复用的多模态流水线架构，以及在生产中平衡延迟、成本与模型质量的工程洞察。 + +### 讲师: + + +
+ +宋辛童：Apache Flink PMC 成员，阿里云高级技术专家 + +宋辛童是 Apache Flink PMC 成员，也是 Flink Agents 子项目的主要推动者。他同时是阿里云的高级技术专家，带领团队从事 Flink 与 AI、湖仓（Lakehouse）融合方向的工作。此前，他在北京大学获得计算机科学博士学位。 \ No newline at end of file diff --git a/content/sessions/dataai-1213409.md b/content/sessions/dataai-1213409.md new file mode 100644 index 000000000..11d43deae --- /dev/null +++ b/content/sessions/dataai-1213409.md @@ -0,0 +1,23 @@ +--- +title: "Building an AI-Powered Spark Diagnosis Assistant Across the Apache Data Stack" +date: "2026-08-07T16:45:00" +track: "dataai" +presenters: "Tianhang Li" +stype: "Chinese Session" +room: "JingMing Hall" +--- + +Modern Spark platforms generate a huge amount of operational signals, but diagnosing performance bottlenecks, resource waste, and recurring production issues still depends heavily on expert experience. In this session, we will share how we built an AI-powered diagnosis assistant that connects multiple Apache big data components, including Gravitino, Spark History Server, YARN ResourceManager, Celeborn, and Ranger, to provide intelligent analysis and actionable recommendations for Spark workloads. + +The assistant helps engineers troubleshoot failed or slow Spark jobs, identify optimization opportunities, and deliver governance insights across performance, resource usage, storage efficiency, and operational stability. By combining metadata, runtime metrics, scheduling signals, shuffle behavior, and access control context, the system can generate practical suggestions for SQL tuning, resource sizing, skew mitigation, shuffle optimization, data governance, and daily on-call troubleshooting. + +We will cover the overall architecture, data collection and reasoning pipeline, real-world diagnosis scenarios, and the measurable impact on compute cost, storage cost, and engineering efficiency. This talk is intended for data platform engineers, Spark practitioners, and open source users interested in applying AI to observability, operations, and optimization in the Apache ecosystem. + +### Speakers: + + +
+ +Tianhang Li: "Big Data Development Engineer at Bilibili | Apache Gravitino Contributor | Expert in Metadata Management & Spark Optimization" + +Li Tianhang is a Big Data Development Engineer at Bilibili, where he specializes in metadata management and Spark computing engine optimization for large-scale data scenarios \ No newline at end of file diff --git a/content/sessions/dataai-1213409.zh.md b/content/sessions/dataai-1213409.zh.md new file mode 100644 index 000000000..7bf6039bb --- /dev/null +++ b/content/sessions/dataai-1213409.zh.md @@ -0,0 +1,23 @@ +--- +title: "跨越 Apache 数据栈，打造 AI 驱动的 Spark 诊断助手" +date: "2026-08-07T16:45:00" +track: "dataai" +presenters: "Tianhang Li" +stype: "中文演讲" +room: "静明厅" +--- + +现代 Spark 平台会产生海量的运行信号，但诊断性能瓶颈、资源浪费和反复出现的生产问题，仍然高度依赖专家经验。在本次演讲中，我们将分享如何构建一个 AI 驱动的诊断助手，它打通了 Gravitino、Spark History Server、YARN ResourceManager、Celeborn 和 Ranger 等多个 Apache 大数据组件，为 Spark 工作负载提供智能分析与可落地的建议。 + +该助手帮助工程师排查失败或缓慢的 Spark 作业、识别优化机会，并在性能、资源使用、存储效率和运维稳定性等维度给出治理洞察。通过整合元数据、运行时指标、调度信号、Shuffle 行为与访问控制上下文，系统能够针对 SQL 调优、资源规格配置、倾斜治理、Shuffle 优化、数据治理以及日常值班排障生成切实可行的建议。 + +我们将介绍整体架构、数据采集与推理流水线、真实诊断场景，以及对计算成本、存储成本和工程效率的可量化影响。本次演讲面向数据平台工程师、Spark 实践者，以及关注如何将 AI 应用于 Apache 生态中可观测性、运维与优化的开源用户。 + +### 讲师: + + +
+ +Tianhang Li："哔哩哔哩大数据开发工程师 | Apache Gravitino 贡献者 | 元数据管理与 Spark 优化专家" + +Li Tianhang 是哔哩哔哩（Bilibili）的大数据开发工程师，专注于大规模数据场景下的元数据管理与 Spark 计算引擎优化。 \ No newline at end of file diff --git a/content/sessions/dataai-1213652.md b/content/sessions/dataai-1213652.md new file mode 100644 index 000000000..655a00a85 --- /dev/null +++ b/content/sessions/dataai-1213652.md @@ -0,0 +1,26 @@ +--- +title: "Ant Real-Time Computing Team's Practice on Flink Expert Agent" +date: "2026-08-07T14:30:00" +track: "dataai" +presenters: "Chaoming Zhang" +stype: "Chinese Session" +room: "JingMing Hall" +--- + +Flink operations and tuning heavily rely on expert experience, with slow anomaly resolution, high remediation costs, and continuous developer queries for SQL optimization and troubleshooting. To address this, we built Flink Ops, an intelligent operations platform integrating self-healing, diagnosis, and Q&A capabilities, and developed Flink Expert Agent powered by LLMs — forming a closed-loop system of "observe-decide-act-interact". +This session covers: +Platform architecture: Unified ingestion of metrics, logs, lineage, and knowledge bases for real-time health profiling +Self-healing engine: Dual-mode automation — rule-based remediation plus Agent-driven intelligent recovery +Intelligent diagnosis: Automated anomaly detection, root-cause reasoning, and remediation recommendation +Expert Q&A: Natural language interface for SQL optimization, configuration tuning, and troubleshooting +Production results: MTTR reduced to X minutes, with Y% developer self-service rate +Who should attend: Data platform engineers, SREs, and AI practitioners interested in intelligent DataOps and LLM-powered operational systems. + +### Speakers: + + +
+ +Chaoming Zhang: Senior Engineer, Ant Real-Time Computing Team + +6 years in big data. Senior Engineer, Ant Flink Engine Team. Building Flink Ops and Expert Agent for intelligent real-time computing operations. \ No newline at end of file diff --git a/content/sessions/dataai-1213652.zh.md b/content/sessions/dataai-1213652.zh.md new file mode 100644 index 000000000..48dbe62fd --- /dev/null +++ b/content/sessions/dataai-1213652.zh.md @@ -0,0 +1,26 @@ +--- +title: "蚂蚁实时计算团队在 Flink Expert Agent 上的实践" +date: "2026-08-07T14:30:00" +track: "dataai" +presenters: "Chaoming Zhang" +stype: "中文演讲" +room: "静明厅" +--- + +Flink 的运维与调优高度依赖专家经验，存在异常处置慢、修复成本高、开发者频繁就 SQL 优化和排障进行咨询等问题。为此，我们打造了集自愈、诊断与问答能力于一体的智能运维平台 Flink Ops，并基于 LLM 开发了 Flink Expert Agent——形成"观测—决策—执行—交互"的闭环体系。 +本次演讲涵盖： +平台架构：统一接入指标、日志、血缘与知识库，实时刻画健康画像 +自愈引擎：双模自动化——基于规则的修复叠加 Agent 驱动的智能恢复 +智能诊断：自动异常检测、根因推理与修复建议 +专家问答：面向 SQL 优化、配置调优与排障的自然语言接口 +生产成效：MTTR 降至 X 分钟，开发者自助率达到 Y% +适用人群：对智能 DataOps 以及 LLM 驱动的运维系统感兴趣的数据平台工程师、SRE 与 AI 从业者。 + +### 讲师: + + +
+ +Chaoming Zhang：高级工程师，蚂蚁实时计算团队 + +6 年大数据经验。蚂蚁 Flink 引擎团队高级工程师。负责构建 Flink Ops 与 Expert Agent，实现实时计算的智能运维。 \ No newline at end of file diff --git a/content/sessions/dataai-1214199.md b/content/sessions/dataai-1214199.md new file mode 100644 index 000000000..a1bad218f --- /dev/null +++ b/content/sessions/dataai-1214199.md @@ -0,0 +1,19 @@ +--- +title: "From Data Lake to Data Agent: Building a Semantic Layer with Apache Gravitino and MetricFlow" +date: "2026-08-08T15:45:00" +track: "dataai" +presenters: "Jerry Shao" +stype: "Chinese Session" +room: "JingMing Hall" +--- + +Natural language data querying is no longer a future promise — it's an engineering problem you can solve today. In this talk, we'll walk through how to combine Apache Gravitino's unified metadata management with dbt's MetricFlow to construct a robust semantic layer that sits between your raw data and your AI agents. You'll learn how Gravitino catalogs and governs your data assets across heterogeneous sources, how MetricFlow defines consistent, reusable metrics on top of them, and how to wire both together so that a data agent can answer business questions reliably — without hallucinating schema details or misinterpreting metric definitions. We'll close with a live demo of an agent asking questions against a multi-source environment and getting accurate, governed answers. + +### Speakers: + + +
+ +Jerry Shao: Datastrato, CTO + +Jerry Shao is the co-founder and CTO of Datastrato, focused on open source Big Data are for more than 10 years. He is an Apache member, committer and PMC member of Apache Spark and Apache Inlong, the original creator of Apache Gravitino. \ No newline at end of file diff --git a/content/sessions/dataai-1214199.zh.md b/content/sessions/dataai-1214199.zh.md new file mode 100644 index 000000000..602aba4e7 --- /dev/null +++ b/content/sessions/dataai-1214199.zh.md @@ -0,0 +1,19 @@ +--- +title: "从数据湖到数据 Agent：用 Apache Gravitino 与 MetricFlow 构建语义层" +date: "2026-08-08T15:45:00" +track: "dataai" +presenters: "Jerry Shao" +stype: "中文演讲" +room: "静明厅" +--- + +自然语言数据查询已不再是未来的愿景——它是你今天就能解决的工程问题。在本次演讲中，我们将 walkthrough 如何把 Apache Gravitino 的统一元数据管理与 dbt 的 MetricFlow 结合起来，在原始数据与 AI Agent 之间构建一个稳健的语义层。你将了解 Gravitino 如何跨异构数据源对数据资产进行编目与治理，MetricFlow 如何在其之上定义一致、可复用的指标，以及如何把两者串联起来，让数据 Agent 能够可靠地回答业务问题——既不会凭空捏造模式细节，也不会误读指标定义。最后，我们将以一个 Agent 在多源环境中提问并获得准确、受治理回答的现场 Demo 收尾。 + +### 讲师: + + +
+ +Jerry Shao：Datastrato CTO + +Jerry Shao 是 Datastrato 联合创始人兼 CTO，专注于开源大数据领域已超过 10 年。他是 Apache 成员、Apache Spark 与 Apache Inlong 的 Committer 及 PMC 成员，也是 Apache Gravitino 的原作者。 \ No newline at end of file diff --git a/content/sessions/dataai-1214460.md b/content/sessions/dataai-1214460.md new file mode 100644 index 000000000..7378932ac --- /dev/null +++ b/content/sessions/dataai-1214460.md @@ -0,0 +1,26 @@ +--- +title: "Architecting the AI-Data Bridge: Exploring the Model Context Protocol (MCP) for Apache Hive" +date: "2026-08-07T15:45:00" +track: "dataai" +presenters: "Attila Turóczy" +stype: "English Session" +room: "JingMing Hall" +--- + +As AI systems evolve, the Model Context Protocol (MCP) is emerging as a standardized way to connect intelligent agents with external systems. Applying this framework to a distributed data warehouse like Apache Hive requires specialized architectural patterns designed for scale and security. This session details the mechanics of exposing Hive through an MCP-compatible interface to build a highly context-aware data source for AI-driven workflows. This session also evaluates the technical feasibility and operational implications of bridging Apache Hive with AI ecosystems using an MCP-driven approach. + +In this session, we will explore the technical mechanics of exposing Apache Hive through an MCP-compatible interface. We will walk through potential architectures that allow AI agents to interact with Hive in a structured, secure, and context-aware manner. By positioning Hive as a first-class data source for AI workflows, capabilities like seamless natural language querying, automated performance analysis, and intelligent, self-serve data exploration can be effectively harnessed. + +This session will contrast existing integration paradigms, including JDBC/ODBC and REST-based query services, with the MCP-centric interaction. Following this evaluation, the session will detail the engineering milestones necessary to build a production-grade MCP server for Hive. Key focal points include establishing dynamic schema awareness, enforcing robust permissioning, implementing query safety guardrails, and optimizing latency for massive data workloads. + +We will conclude with a pragmatic cost-benefit analysis of deploying an MCP server for Hive. By weighing the immediate business value against the operational complexities introduced, we will provide a framework to determine whether building a MCP server for Hive represents a worthwhile engineering investment for modern data teams. + +This talk is for data engineers, system architects, and open-source contributors who are exploring the intersection of AI agents and data platforms, and want a realistic view of what's possible with Hive today. + +### Speakers: + +
+ +Attila Turóczy: Senior Director of Engineering at Cloudera + +Apache Hive, Impala and Big Data enthusiasm at Cloudera \ No newline at end of file diff --git a/content/sessions/dataai-1214460.zh.md b/content/sessions/dataai-1214460.zh.md new file mode 100644 index 000000000..cd62357f8 --- /dev/null +++ b/content/sessions/dataai-1214460.zh.md @@ -0,0 +1,26 @@ +--- +title: "架设 AI 与数据的桥梁：探索面向 Apache Hive 的模型上下文协议（MCP）" +date: "2026-08-07T15:45:00" +track: "dataai" +presenters: "Attila Turóczy" +stype: "英文演讲" +room: "静明厅" +--- + +随着 AI 系统不断演进，模型上下文协议（Model Context Protocol，MCP）正成为一种连接智能体与外部系统的标准化方式。将这一框架应用到 Apache Hive 这样的分布式数据仓库上，需要专门面向规模化与安全性而设计的架构模式。本次演讲将详细讲解如何通过 MCP 兼容接口把 Hive 暴露出来，为 AI 驱动的工作流构建一个高度上下文感知的数据源。同时，我们也会评估用 MCP 驱动的方式打通 Apache Hive 与 AI 生态的技术可行性与运维影响。 + +在本次演讲中，我们将探讨通过 MCP 兼容接口暴露 Apache Hive 的技术机理，梳理出若干让 AI Agent 能够以结构化、安全且上下文感知的方式与 Hive 交互的潜在架构。把 Hive 定位为 AI 工作流的一等数据源后，无缝的自然语言查询、自动化性能分析，以及智能的自助式数据探索等能力便可被有效释放出来。 + +本次演讲会将既有的集成范式——包括 JDBC/ODBC 与基于 REST 的查询服务——与 MCP 为中心的交互方式进行对比。在此评估之后，我们将详细说明为 Hive 构建生产级 MCP 服务器所需的工程里程碑。重点关注的方向包括：建立动态的模式感知、强化权限管控、实现查询安全护栏，以及为海量数据工作负载优化延迟。 + +最后，我们将对为 Hive 部署 MCP 服务器进行务实的成本效益分析。通过权衡即时可得的业务价值与由此引入的运维复杂度，我们将提供一个判断框架，帮助判断为 Hive 构建 MCP 服务器对现代数据团队而言是否是一项值得的工程投入。 + +本次演讲面向正在探索 AI Agent 与数据平台交汇点、并希望了解 Hive 当前能做到什么程度的数据工程师、系统架构师与开源贡献者。 + +### 讲师: + +
+ +Attila Turóczy：Cloudera 工程高级总监 + +在 Cloudera 致力于 Apache Hive、Impala 及大数据领域的推广。 \ No newline at end of file diff --git a/content/sessions/dataai-1300514.md b/content/sessions/dataai-1300514.md new file mode 100644 index 000000000..f4c04281e --- /dev/null +++ b/content/sessions/dataai-1300514.md @@ -0,0 +1,19 @@ +--- +title: "Who Pays for the Data? Metering Apache Pipelines for AI Agents" +date: "2026-08-08T16:45:00" +track: "dataai" +presenters: "Mike Ma" +stype: "Chinese Session" +room: "JingMing Hall" +--- + +Solana's data infrastructure already runs on Apache — Kafka, Parquet, Arrow, DataFusion, Airflow, Beam — used by Coinbase, Allium, Blockdaemon and other ecosystem teams. What's missing from the stack is metering: nothing today answers "who pays for this call, how much, and to whom?" This talk walks through how the busiest open-source streaming pipelines on Solana are empowered by the Apache stack, proposes adding an x402 (HTTP 402 Payment Required) middleware at the data-plane gateway so open-source projects can charge AI agents per call, and reviews real success cases of onchain data monetization. + +### Speakers: + + +
+ +Mike Ma: Developer Advocate @Solana Foundation + +Mike Ma is developer advocate at Solana Foundation and an open source developer in apache ecosystem and Solana ecosystem contributing Solana-go SDK, x402 protocol, APP agentic payment protocol and initiated Solar - the chinese speaking Solana developer community. \ No newline at end of file diff --git a/content/sessions/dataai-1300514.zh.md b/content/sessions/dataai-1300514.zh.md new file mode 100644 index 000000000..283741447 --- /dev/null +++ b/content/sessions/dataai-1300514.zh.md @@ -0,0 +1,19 @@ +--- +title: "谁为数据买单？为 AI Agent 计量 Apache 数据管道" +date: "2026-08-08T16:45:00" +track: "dataai" +presenters: "Mike Ma" +stype: "中文演讲" +room: "静明厅" +--- + +Solana 的数据基础设施早已运行在 Apache 之上——Kafka、Parquet、Arrow、DataFusion、Airflow、Beam 已被 Coinbase、Allium、Blockdaemon 等生态团队广泛采用。而 Apache 栈中缺失的一环是"计量"：目前没有任何组件能回答"这次调用谁来付、付多少、付给谁？"。本次分享将呈现 Solana 上最繁忙的开源数据管道如何依托 Apache 栈，并提出在数据面网关引入 x402（HTTP 402 Payment Required）中间件的方案，让开源项目可以对 AI Agent 按次计费；同时回顾链上数据变现的真实成功案例。 + +### 讲师: + + +
+ +Mike Ma：Solana Foundation 开发者布道师 + +Mike Ma 现任 Solana Foundation 开发者布道师，同时也是 Apache 生态与 Solana 生态的开源开发者，贡献了 Solana-go SDK、x402 协议、APP（Agentic Payment Protocol，智能体支付协议），并创办了 Solar——中文 Solana 开发者社区。 \ No newline at end of file diff --git a/content/sessions/datalake-1206028.md b/content/sessions/datalake-1206028.md new file mode 100644 index 000000000..7616a0e42 --- /dev/null +++ b/content/sessions/datalake-1206028.md @@ -0,0 +1,19 @@ +--- +title: "Iceberg UDF Spec: Portable SQL Functions Across Engines" +date: "2026-08-09T13:30:00" +track: "datalake" +presenters: "Huaxin Gao" +stype: "English Session" +room: "MainRoom - YiHe Hall" +--- + +Iceberg’s UDF spec introduces a self-contained, versioned metadata format for SQL scalar and table functions that can move across catalogs and engines like Spark and Trino. This talk walks through the core model, definitions, parameter naming rules, return types, and dialect-specific representations, and explains how versioning, rollback, determinism, and null-handling hints make UDFs portable without sacrificing engine optimizations. We’ll highlight key design choices such as immutable signatures, overload compatibility with defaults, and secure functions, and close with practical guidance for implementing the spec in an engine or catalog. + +### Speakers: + + +
+ +Huaxin Gao: Software engineer at Snowflake + +Huaxin Gao is a software engineer at Snowflake and an Apache Spark committer and PMC member. She is also a committer for Apache Iceberg and Apache DataFusion Comet, with contributions spanning query engines, table formats, and distributed data systems. \ No newline at end of file diff --git a/content/sessions/datalake-1206028.zh.md b/content/sessions/datalake-1206028.zh.md new file mode 100644 index 000000000..6ca7795c2 --- /dev/null +++ b/content/sessions/datalake-1206028.zh.md @@ -0,0 +1,19 @@ +--- +title: "Iceberg UDF 规范：跨引擎的可移植 SQL 函数" +date: "2026-08-09T13:30:00" +track: "datalake" +presenters: "Huaxin Gao" +stype: "英文演讲" +room: "主会场 - 颐和厅" +--- + +Iceberg 的 UDF 规范引入了一种自包含、带版本的元数据格式，用于 SQL 标量函数和表函数，它们可以跨 catalog 和引擎（如 Spark 和 Trino）流转。本次演讲将走查核心模型、定义、参数命名规则、返回类型以及方言特定的表示方式，并解释版本管理、回滚、确定性（determinism）和空值处理提示如何让 UDF 具备可移植性，同时不牺牲引擎侧的优化。我们将重点介绍一些关键设计选择，例如不可变的函数签名、与默认值兼容的重载，以及安全函数（secure function），最后给出在引擎或 catalog 中实现该规范的实用指引。 + +### 讲师: + + +
+ +Huaxin Gao：Snowflake 软件工程师 + +Huaxin Gao 是 Snowflake 的软件工程师，也是 Apache Spark 的 committer 和 PMC 成员。她还是 Apache Iceberg 和 Apache DataFusion Comet 的 committer，贡献横跨查询引擎、表格式和分布式数据系统。 \ No newline at end of file diff --git a/content/sessions/datalake-1207558.md b/content/sessions/datalake-1207558.md new file mode 100644 index 000000000..869c4a199 --- /dev/null +++ b/content/sessions/datalake-1207558.md @@ -0,0 +1,21 @@ +--- +title: "Building Audit and Integrations with Apache Polaris Events" +date: "2026-08-07T16:15:00" +track: "datalake" +presenters: "Adnan Hemani" +stype: "English Session" +room: "MainRoom - YiHe Hall" +--- + +Apache Iceberg REST Catalogs are rapidly becoming the standard for interoperable data catalogs, and Apache Polaris (Incubating) is an Apache project implementation of the Apache Iceberg REST API. One of the things that makes Polaris so powerful is how it lets you observe and react to changes in your data ecosystem. + +This session will introduce Apache Polaris, and the Polaris Events functionality to a beginner-level audience. We will demystify how the platform tracks critical activity and explore why these events matter for practical needs like data auditing, compliance, monitoring, integrations, and automation. To prove just how accessible the platform is, the session features a live coding demonstration where we will build and deploy a custom customer event listener in just a few minutes. Finally, we'll show upcoming planned improvements - and where you can help! + +### Speakers: + + +
+ +Adnan Hemani: Senior Software Engineer at Snowflake + +Adnan Hemani is currently a Senior Software Engineer at Snowflake, where he currently works on open source technologies, such as Apache Iceberg and Apache Polaris. Previously, he led multiple teams within the Amazon Web Services (AWS) EMR (ElasticMapReduce) organization with a special focus on Data Access Controls across all Big Data engines that are offered on the product (Spark, Trino, Presto, Hive, etc.) \ No newline at end of file diff --git a/content/sessions/datalake-1207558.zh.md b/content/sessions/datalake-1207558.zh.md new file mode 100644 index 000000000..1b7e24796 --- /dev/null +++ b/content/sessions/datalake-1207558.zh.md @@ -0,0 +1,21 @@ +--- +title: "基于 Apache Polaris Events 构建审计与集成" +date: "2026-08-07T16:15:00" +track: "datalake" +presenters: "Adnan Hemani" +stype: "英文演讲" +room: "主会场 - 颐和厅" +--- + +Apache Iceberg REST Catalog 正迅速成为可互操作数据 catalog 的事实标准，而 Apache Polaris（孵化中）是 Apache Iceberg REST API 的一个 Apache 项目实现。让 Polaris 如此强大的原因之一，在于它让你能够观察并对数据生态系统中的变化做出反应。 + +本次演讲将面向入门级听众介绍 Apache Polaris 及其 Polaris Events 功能。我们将揭开该平台如何追踪关键活动的面纱，并探讨这些事件为何对数据审计、合规、监控、集成和自动化等实际需求如此重要。为了证明该平台有多易上手，演讲还包含一场现场编程演示——我们将在几分钟内构建并部署一个自定义的客户事件监听器。最后，我们会介绍即将到来的计划改进——以及你可以在哪些方面贡献力量！ + +### 讲师: + + +
+ +Adnan Hemani：Snowflake 高级软件工程师 + +Adnan Hemani 目前是 Snowflake 的高级软件工程师，从事 Apache Iceberg 和 Apache Polaris 等开源技术方面的工作。此前，他曾在 Amazon Web Services（AWS）EMR（ElasticMapReduce）组织中带领多个团队，特别专注于该产品所提供的所有大数据引擎（Spark、Trino、Presto、Hive 等）的数据访问控制。 \ No newline at end of file diff --git a/content/sessions/datalake-1208879.md b/content/sessions/datalake-1208879.md new file mode 100644 index 000000000..04b8e7638 --- /dev/null +++ b/content/sessions/datalake-1208879.md @@ -0,0 +1,23 @@ +--- +title: "Beyond Metadata: Advanced Iceberg Management with Apache Gravitino" +date: "2026-08-07T14:00:00" +track: "datalake" +presenters: "Xiaojing Fang" +stype: "Chinese Session" +room: "MainRoom - YiHe Hall" +--- + +Apache Iceberg provides a strong foundation for open data lake tables, but production-grade deployments often need more than a table format. They need unified metadata access, secure authorization, cloud-native credential management, query acceleration, and automated maintenance. + +In this session, we will present how Apache Gravitino enhances Iceberg with advanced management and acceleration capabilities. We will show how Gravitino provides a unified metadata view, enables federated analytics across Iceberg and Lance, supports credential vending in cloud environments, and offers centralized authorization for multiple engines. We will also cover performance-oriented features such as scan planning and table cache, as well as the Table Maintenance Service for automated table operations. + +This session will be useful for engineers and architects building secure, open, and high-performance lakehouse platforms on top of Apache Iceberg. + +### Speakers: + + +
+ +Xiaojing Fang: Apache Gravitino Committer + +Apache Gravitino Committer, architect at China Mobile, focusing on data and AI infrastructure. \ No newline at end of file diff --git a/content/sessions/datalake-1208879.zh.md b/content/sessions/datalake-1208879.zh.md new file mode 100644 index 000000000..265d70e08 --- /dev/null +++ b/content/sessions/datalake-1208879.zh.md @@ -0,0 +1,23 @@ +--- +title: "超越元数据：用 Apache Gravitino 进行进阶 Iceberg 管理" +date: "2026-08-07T14:00:00" +track: "datalake" +presenters: "Xiaojing Fang" +stype: "中文演讲" +room: "主会场 - 颐和厅" +--- + +Apache Iceberg 为开放数据湖表提供了坚实的基础，但生产级部署往往需要的不仅是一个表格式。它们还需要统一的元数据访问、安全的授权、云原生的凭据管理、查询加速，以及自动化的维护。 + +在本次演讲中，我们将介绍 Apache Gravitino 如何通过高级管理与加速能力增强 Iceberg。我们将展示 Gravitino 如何提供统一的元数据视图、支持跨 Iceberg 和 Lance 的联邦分析、在云环境中支持 credential vending，并为多个引擎提供集中化授权。我们还会涵盖面向性能的特性，如 scan planning 和 table cache，以及用于自动化表操作的 Table Maintenance Service。 + +本次演讲对那些基于 Apache Iceberg 构建安全、开放、高性能湖仓平台的工程师和架构师会有所帮助。 + +### 讲师: + + +
+ +Xiaojing Fang：Apache Gravitino Committer + +Apache Gravitino Committer，中国移动架构师，专注于数据与 AI 基础设施。 \ No newline at end of file diff --git a/content/sessions/datalake-1209515.md b/content/sessions/datalake-1209515.md new file mode 100644 index 000000000..abba3830e --- /dev/null +++ b/content/sessions/datalake-1209515.md @@ -0,0 +1,38 @@ +--- +title: "Fluss X Lakehouse: Unlocking Sub-Second Data Freshness on Your Data Lake" +date: "2026-08-08T16:15:00" +track: "datalake" +presenters: "Yuxia Luo" +stype: "Chinese Session" +room: "MainRoom - YiHe Hall" +--- + +Modern data platforms are torn between two worlds: streaming systems deliver fresh data but lack long-term + storage, while lakehouses provide durable analytics but can't serve truly real-time queries. For primary-key + tables, this gap is especially painful — Paimon requires compaction before new rows become visible, and Iceberg + accumulates equality delete files that cripple query performance. + + Apache Fluss (incubating) closes this gap as a dedicated streaming storage layer that unifies with the lakehouse. + In this talk, we'll walk through: + + - The Tiering Service that continuously syncs data from Fluss to Paimon, Iceberg, and Hudi + - Union Read, which transparently merges real-time Fluss data with historical lake data in a single query + - How Fluss leverages its native primary-key index to generate Deletion Vectors during tiering, eliminating the + compaction bottleneck in Paimon and the equality-delete overhead in Iceberg + - A live end-to-end demo using Fluss + Iceberg/Paimon + DuckDB, showing sub-second freshness on primary-key tables + + We'll close with the roadmap: multi-engine Union Read (Spark, Trino, StarRocks), broader lake support (Hudi, + Delta), and heterogeneous tiering within a single Fluss cluster. + +### Speakers: + + +
+ +Yuxia Luo: Apache Fluss PMC | Software Engineer @ Alibaba + +Yuxia Luo is a software engineer at Alibaba and a PMC member of Apache Fluss, working on streaming + storage and its convergence with modern lakehouse formats. He drives features around Tiering, Union Read, and + real-time queryability for primary-key tables across Paimon, Iceberg. With years of experience on Apache + Flink and large-scale real-time data platforms, he is passionate about closing the gap between streaming and the + data lake. He has previously shared his work at Flink Forward and QCon. \ No newline at end of file diff --git a/content/sessions/datalake-1209515.zh.md b/content/sessions/datalake-1209515.zh.md new file mode 100644 index 000000000..7eed7f998 --- /dev/null +++ b/content/sessions/datalake-1209515.zh.md @@ -0,0 +1,29 @@ +--- +title: "Fluss × 湖仓：在数据湖上解锁亚秒级数据新鲜度" +date: "2026-08-08T16:15:00" +track: "datalake" +presenters: "Yuxia Luo" +stype: "中文演讲" +room: "主会场 - 颐和厅" +--- + +现代数据平台被夹在两个世界之间：流式系统交付新鲜数据，却缺乏长期存储；湖仓提供持久的分析能力，却无法服务真正的实时查询。对于主键表而言，这一鸿沟尤为痛苦——Paimon 需要经过 compaction 才能让新行可见，而 Iceberg 则会累积 equality delete 文件，从而拖垮查询性能。 + + Apache Fluss（孵化中）作为一个与湖仓统一的专用流式存储层，弥合了这一鸿沟。 + 在本次演讲中，我们将走查： + + - Tiering Service：持续地把数据从 Fluss 同步到 Paimon、Iceberg 和 Hudi + - Union Read：在一次查询中透明地合并实时 Fluss 数据与历史湖数据 + - Fluss 如何利用其原生主键索引，在分层（tiering）过程中生成 Deletion Vector，从而消除 Paimon 的 compaction 瓶颈和 Iceberg 的 equality-delete 开销 + - 一场使用 Fluss + Iceberg/Paimon + DuckDB 的端到端现场演示，展示主键表上的亚秒级新鲜度 + + 我们将以路线图收尾：多引擎 Union Read（Spark、Trino、StarRocks）、更广泛的湖支持（Hudi、Delta），以及单个 Fluss 集群内的异构分层（tiering）。 + +### 讲师: + + +
+ +Yuxia Luo：Apache Fluss PMC | 阿里巴巴软件工程师 + +Yuxia Luo 是阿里巴巴的软件工程师，也是 Apache Fluss 的 PMC 成员，从事流式存储及其与现代湖仓格式融合的工作。他主导了围绕 Tiering、Union Read 以及跨 Paimon、Iceberg 的主键表实时可查询性等特性。凭借在 Apache Flink 和大规模实时数据平台方面多年的经验，他热衷于弥合流式与数据湖之间的鸿沟。他曾在此前的 Flink Forward 和 QCon 上分享过他的工作。 \ No newline at end of file diff --git a/content/sessions/datalake-1210488.md b/content/sessions/datalake-1210488.md new file mode 100644 index 000000000..d1794405c --- /dev/null +++ b/content/sessions/datalake-1210488.md @@ -0,0 +1,23 @@ +--- +title: "Lakehouse to AI Data Lake: Xiaomi AI scene data storage and computing architecture evolution" +date: "2026-08-09T14:00:00" +track: "datalake" +presenters: "Kainan Bao" +stype: "Chinese Session" +room: "MainRoom - YiHe Hall" +--- + +While the lakehouse architecture has become the de facto standard for business intelligence (BI) workloads, it faces two fundamental bottlenecks when extended to large-scale foundation model training pipelines. Unlike traditional BI which centers on structured tabular data, AI training processes massive volumes of unstructured data (text, images, audio, video), exposing critical gaps in unified unstructured data governance and architectural silos between Hadoop-based data processing and cloud-native model training infrastructure. + +This session presents Xiaomi's production-proven two-phase architecture evolution to address these challenges. Phase 1 integrates unstructured data into a unified metadata governance layer using Gravitino and Fileset, and enables seamless interoperability between cloud object storage and the Hadoop ecosystem via GVFS. Phase 2 adopts a Ray-Lance stack that transcends the limitations of the traditional Hadoop stack, leveraging Ray as the cloud-native distributed computing engine and Lance as the AI-optimized storage format. Attendees will learn practical implementation details and real-world performance gains from production use cases including incremental data deduplication and multimodal data management. + +### Speakers: + + +
+ +Kainan Bao: Software Engineer at Xiaomi | Core Contributor of Xiaomi Mimo Models | Contributor of Iceberg, Paimon & Gravitino + +Software Engineer at Xiaomi and a Core Contributor of Xiaomi Mimo Models, responsible for PB-scale data processing for Xiaomi Mimo foundation model training. Specializes in solving the unique challenges of processing massive volumes of multimodal unstructured data including text, images and video. + +Has extensive experience in data lake governance, with contributions to the Iceberg, Paimon & Gravitino open source communities. In this session, will share how Xiaomi evolved its traditional data lake to unify structured and unstructured data governance with Gravitino, and built AI-native data processing systems using Ray and Lance. \ No newline at end of file diff --git a/content/sessions/datalake-1210488.zh.md b/content/sessions/datalake-1210488.zh.md new file mode 100644 index 000000000..f0b522c2a --- /dev/null +++ b/content/sessions/datalake-1210488.zh.md @@ -0,0 +1,23 @@ +--- +title: "从湖仓到 AI 数据湖：小米 AI 场景数据存储与计算架构演进" +date: "2026-08-09T14:00:00" +track: "datalake" +presenters: "Kainan Bao" +stype: "中文演讲" +room: "主会场 - 颐和厅" +--- + +尽管湖仓架构已成为商业智能（BI）工作负载的事实标准，但当它被扩展到大规模基础模型训练管道时，却面临两个根本性瓶颈。与以结构化表格数据为中心的传统 BI 不同，AI 训练需要处理海量的非结构化数据（文本、图像、音频、视频），这暴露出统一非结构化数据治理方面的关键缺口，以及基于 Hadoop 的数据处理与云原生模型训练基础设施之间的架构孤岛。 + +本次演讲介绍小米经过生产验证的两阶段架构演进，以应对这些挑战。第一阶段使用 Gravitino 和 Fileset 将非结构化数据纳入统一的元数据治理层，并通过 GVFS 实现云对象存储与 Hadoop 生态之间的无缝互操作。第二阶段采用 Ray-Lance 技术栈，突破了传统 Hadoop 技术栈的局限——以 Ray 作为云原生分布式计算引擎，以 Lance 作为面向 AI 优化的存储格式。听众将了解到实际的实现细节，以及来自生产用例（包括增量数据去重和多模态数据管理）的真实性能收益。 + +### 讲师: + + +
+ +Kainan Bao：小米软件工程师 | 小米 Mimo 模型核心贡献者 | Iceberg、Paimon 与 Gravitino 的贡献者 + +小米软件工程师，小米 Mimo 模型核心贡献者，负责小米 Mimo 基础模型训练的 PB 级数据处理。擅长解决处理海量多模态非结构化数据（包括文本、图像和视频）的独特挑战。 + +在数据湖治理方面经验丰富，曾向 Iceberg、Paimon 与 Gravitino 开源社区贡献。在本次演讲中，他将分享小米如何演进其传统数据湖、用 Gravitino 统一结构化与非结构化数据治理，并使用 Ray 和 Lance 构建 AI 原生数据处理系统。 \ No newline at end of file diff --git a/content/sessions/datalake-1211108.md b/content/sessions/datalake-1211108.md new file mode 100644 index 000000000..f02d66ec8 --- /dev/null +++ b/content/sessions/datalake-1211108.md @@ -0,0 +1,26 @@ +--- +title: "Apache Iceberg V3 in Production: Lessons from a Large-Scale Deployment" +date: "2026-08-08T14:00:00" +track: "datalake" +presenters: "Yuming Wang, Fei Wang" +stype: "Chinese Session" +room: "MainRoom - YiHe Hall" +--- + +Apache Iceberg V3 brings powerful capabilities, but running it reliably in production takes more than a version upgrade. In this session, we share our real-world adoption journey with Iceberg 1.10: table design, write/read path tuning, compaction strategy, and metadata lifecycle management. We’ll unpack the key bottlenecks we faced—small-file growth, snapshot sprawl, merge/update pressure, and query latency variance—and the concrete optimizations that improved both stability and cost efficiency. You’ll leave with practical architecture patterns, anti-patterns, and a battle-tested checklist to move from pilot to production with confidence. + +### Speakers: + + +
+ +Yuming Wang: eBay, Compute, Lakehouse + +Yuming Wang is a member of the Apache Spark Project Management Committee (PMC) and currently leads the build-out and evolution of his company’s compute platform. He focuses on architecting, optimizing, and operating Spark at production scale, while also driving the implementation and operation of a Lakehouse architecture based on Apache Iceberg. He is committed to turning cutting-edge open-source data technologies into reliable and efficient enterprise-grade production systems. + + + +
+Fei Wang: eBay, Hadoop + +Fei Wang is a PMC member of both Apache Kyuubi and Apache Celeborn, currently focusing on the development and evolution of enterprise computing platforms. He specializes in distributed computing, Spark engine optimization, and Lakehouse architecture based on Apache Iceberg, dedicated to translating open-source data technologies into stable and efficient enterprise-grade computing services. \ No newline at end of file diff --git a/content/sessions/datalake-1211108.zh.md b/content/sessions/datalake-1211108.zh.md new file mode 100644 index 000000000..d93229aeb --- /dev/null +++ b/content/sessions/datalake-1211108.zh.md @@ -0,0 +1,25 @@ +--- +title: "生产环境中的 Apache Iceberg V3：来自大规模部署的经验教训" +date: "2026-08-08T14:00:00" +track: "datalake" +presenters: "Yuming Wang, Fei Wang" +stype: "中文演讲" +room: "主会场 - 颐和厅" +--- + +Apache Iceberg V3 带来了强大的能力，但要可靠地在生产环境中运行它，需要的远不止一次版本升级。在本次演讲中，我们分享采用 Iceberg 1.10 的真实历程：表设计、写入/读取路径调优、compaction 策略，以及元数据生命周期管理。我们将剖析所面临的关键瓶颈——小文件增长、快照膨胀、合并/更新压力，以及查询延迟波动——以及那些既提升了稳定性又提高了成本效率的具体优化。你将带走实用的架构模式、反模式，以及一份身经百战的检查清单，从而有信心地从试点走向生产。 + +### 讲师: + + +
+ +Yuming Wang：eBay，计算与湖仓 + +Yuming Wang 是 Apache Spark 项目管理委员会（PMC）的成员，目前负责公司计算平台的建设与演进。他专注于在生产规模下对 Spark 进行架构设计、优化和运营，同时推动基于 Apache Iceberg 的湖仓架构的落地与运营。他致力于把前沿的开源数据技术转化为可靠、高效的企业级生产系统。 + + +
+Fei Wang：eBay，Hadoop + +Fei Wang 是 Apache Kyuubi 与 Apache Celeborn 项目管理委员会（PMC）成员，目前参与公司计算平台的建设与演进。他专注于分布式计算、Spark 引擎优化以及基于 Apache Iceberg 的 Lakehouse 架构实践，致力于将开源数据技术转化为稳定高效的企业级计算服务。 \ No newline at end of file diff --git a/content/sessions/datalake-1211745.md b/content/sessions/datalake-1211745.md new file mode 100644 index 000000000..d564395c3 --- /dev/null +++ b/content/sessions/datalake-1211745.md @@ -0,0 +1,39 @@ +--- +title: "Evolving a real-time lakehouse: Stability and performance breakthroughs at scale" +date: "2026-08-08T15:45:00" +track: "datalake" +presenters: "Zhuojun Jiang, Wenling Zhang" +stype: "Chinese Session" +room: "MainRoom - YiHe Hall" +--- + +As real-time data processing becomes the foundation of modern data platforms, the challenge is no longer just enabling real-time ingestion, but ensuring stability, consistency, and high performance under large-scale and high-concurrency workloads. Following our previous practice on building a real-time lakehouse, we further evolved the architecture to address critical bottlenecks in large-scale production environments. + +In this session, we present a production-grade real-time lakehouse architecture based on Flink CDC 3.4, Apache Iceberg, and Apache Amoro, designed to support large-scale multi-business and highly sharded tables with second-level ingestion latency. + +We redesigned the schema evolution mechanism to address performance bottlenecks and risks caused by redundant downstream writes during schema changes, achieving over 80% improvement in write efficiency. At the same time, we resolved schema consistency issues in distributed environments, significantly improving synchronization stability under high concurrency. + +On the optimization side, by leveraging Apache Amoro’s adaptive refresh and zero-copy compaction mechanisms, we reduced Metastore load while achieving up to 4× improvement in file compaction performance. In addition, system robustness was enhanced through improvements in service restart and operational stability. + +This session will cover two key areas: + +- Distributed schema evolution and consistency guarantees for large-scale real-time ingestion +- Adaptive optimization techniques for improving performance and stability in lakehouse systems + +These practices have been validated in production at telecom-scale workloads, providing a practical approach to building a high-performance and stable real-time lakehouse. + +### Speakers: + + +
+ +Zhuojun Jiang: Senior Big Data Engineer, State Cloud + +Zhuojun Jiang is a Senior Big Data Engineer at State Cloud, specializing in real-time data lakehouse architecture. She focuses on big data development, system performance optimization, and real-time data synchronization based on technologies like Apache FlinkCDC, Iceberg, and Amoro. Jiang actively contributes to the open-source community and shares practical insights through industry forums and technical talks. + + +
+ +Wenling Zhang: Senior Big Data Engineer, State Cloud + +Apache (incubating) Amoro contributor \ No newline at end of file diff --git a/content/sessions/datalake-1211745.zh.md b/content/sessions/datalake-1211745.zh.md new file mode 100644 index 000000000..2e6674a2e --- /dev/null +++ b/content/sessions/datalake-1211745.zh.md @@ -0,0 +1,38 @@ +--- +title: "实时湖仓的演进：面向大规模场景的稳定性与性能突破" +date: "2026-08-08T15:45:00" +track: "datalake" +presenters: "Zhuojun Jiang, Wenling Zhang" +stype: "中文演讲" +room: "主会场 - 颐和厅" +--- + +随着实时数据平台规模不断扩大，在高并发与大规模数据场景下保障系统的稳定性、一致性与高性能，成为实时湖仓建设的核心挑战。我们基于之前的实时湖仓实践，进一步对架构进行了持续演进，重点解决生产环境中的关键瓶颈问题。 + +本次分享将介绍一个基于 Flink CDC 3.4、Apache Iceberg 与 Apache Amoro 构建的生产级实时湖仓架构，支持多业务表、多分片场景下的秒级数据入湖能力。 + +在数据同步层，我们重构了 Schema Evolution 架构，消除了模式变更过程中下游重复写入带来的性能瓶颈，使写入效率提升 80% 以上。同时，针对分布式环境中的模式一致性问题进行了系统性优化，显著提升了高并发场景下的数据同步稳定性。 + +在数据优化层，基于 Apache Amoro 的自适应刷新与零拷贝合并机制，在显著降低 Metastore 负载的同时，实现了 4 倍的文件合并性能提升，并通过优化服务重启与运行机制进一步增强系统整体稳定性。 + +本次分享将重点介绍两方面实践经验： +- 面向大规模实时同步的分布式 Schema 演进与一致性保障 +- 基于自适应优化机制的湖仓性能优化与系统稳定性提升 + +相关能力已在电信运营商生产环境中落地，支撑 PB 级数据规模，为构建高性能、高稳定性的实时湖仓提供可落地的实践路径。 + +### 讲师: + + +
+ +Zhuojun Jiang：天翼云 高级大数据工程师 + +Zhuojun Jiang 是电信天翼云的一名大数据开发工程师，也是Apache Amoro (incubating) 项目的Committer。她正致力于实时大数据湖仓架构研发与优化，专注于基于Apache FlinkCDC、Iceberg 和 Amoro 等组件搭建的大数据湖仓设计与开发，并积极参与开源社区贡献。 + + +
+ +Wenling Zhang：天翼云 高级大数据工程师 + +Wenling Zhang 是天翼云的一名大数据开发工程师，也是 Apache Amoro (incubating) 的 contributor。目前主要从事 Apache Iceberg、Amoro、Lance 等数据湖开源组件的开发与性能优化工作。 \ No newline at end of file diff --git a/content/sessions/datalake-1212302.md b/content/sessions/datalake-1212302.md new file mode 100644 index 000000000..46493dc55 --- /dev/null +++ b/content/sessions/datalake-1212302.md @@ -0,0 +1,24 @@ +--- +title: "Lean Metadata, Big Data: Keep the Query Fast in Apache Iceberg" +date: "2026-08-09T14:30:00" +track: "datalake" +presenters: "Hongyue Zhang" +stype: "English Session" +room: "MainRoom - YiHe Hall" +--- + +Apache Iceberg's metadata layer is what makes schema evolution, partition evolution, and predicate pushdown possible . But at petabyte scale, that metadata itself can balloon to hundreds of gigabytes or more. When planning slows down the execution, something has to give. + +This talk explores how Iceberg's metadata is structured through manifests, partition summaries, and column-level metrics, and why each layer exists to enable fast analytical queries. We'll then walk through Iceberg's native APIs, procedures and practical strategies for applying them to keep metadata size in check at scale. Finally, we'll preview proposals in the Iceberg v4 spec that aim to make metadata more compact and scalable by default. + +You'll walk away with a clear picture of where all those gigabytes of metadata actually come from, a playbook for keeping them in check in an iceberg's own way, and enough context on the v4 spec to start contributing and where the format heads next. + + +### Speakers: + + +
+ +Hongyue Zhang: Software Engineer at Snowflake + +Hongyue started to contribute to apache iceberg project since 2022 while work on Apple data platform. Now at Snowflake, he is building tools and systems around Apache Iceberg to help make data-driven decisions. \ No newline at end of file diff --git a/content/sessions/datalake-1212302.zh.md b/content/sessions/datalake-1212302.zh.md new file mode 100644 index 000000000..64024c36f --- /dev/null +++ b/content/sessions/datalake-1212302.zh.md @@ -0,0 +1,24 @@ +--- +title: "精简元数据，驾驭大数据：让 Apache Iceberg 中的查询保持飞快" +date: "2026-08-09T14:30:00" +track: "datalake" +presenters: "Hongyue Zhang" +stype: "英文演讲" +room: "主会场 - 颐和厅" +--- + +Apache Iceberg 的元数据层正是 schema 演进、分区演进和谓词下推（predicate pushdown）得以实现的基础。但在 PB 级规模下，这些元数据本身可能膨胀到数百 GB 甚至更多。当规划阶段拖慢了执行阶段，就总得有所取舍。 + +本次演讲探讨 Iceberg 的元数据是如何通过 manifest、分区摘要（partition summary）以及列级指标来组织的，以及为什么每一层都是为了支撑快速分析查询而存在。随后，我们将走查 Iceberg 的原生 API、procedure，以及把它们应用于大规模场景下控制元数据体量的实用策略。最后，我们会预览 Iceberg v4 规范中的一些提案，它们旨在让元数据在默认情况下更紧凑、更具可扩展性。 + +你将清晰地了解那数 GB 的元数据究竟从何而来，获得一份以 Iceberg 自身方式来控制元数据体量的"打法手册"，以及对 v4 规范足够的背景认知——足以让你开始参与贡献，并了解该格式未来的走向。 + + +### 讲师: + + +
+ +Hongyue Zhang：Snowflake 软件工程师 + +Hongyue 自 2022 年起开始为 Apache Iceberg 项目做贡献，当时他在 Apple 数据平台工作。如今在 Snowflake，他围绕 Apache Iceberg 构建工具和系统，以帮助做出数据驱动的决策。 \ No newline at end of file diff --git a/content/sessions/datalake-1212404.md b/content/sessions/datalake-1212404.md new file mode 100644 index 000000000..dda0f40e4 --- /dev/null +++ b/content/sessions/datalake-1212404.md @@ -0,0 +1,28 @@ +--- +title: "Building a Unified Lakehouse: Best Practices with Apache Paimon and Ecosystem" +date: "2026-08-07T14:30:00" +track: "datalake" +presenters: "Zhoulong Liu" +stype: "Chinese Session" +room: "MainRoom - YiHe Hall" +--- + +As data architectures evolve, maintaining separate silos for batch and streaming processing — the hallmark of Lambda architecture — has become increasingly costly and complex. How can we build a truly unified platform that delivers both real-time data freshness and high-performance analytics at scale? + +In this talk, we go beyond theory and dive into the trenches of building a next-generation Unified Lakehouse centered around Apache Paimon. We will share battle-tested best practices and real-world implementation patterns, demonstrating how to architect a seamless data pipeline through deep ecosystem synergy: + +Apache Flink + Paimon — Robust, low-latency real-time ingestion into a transactional data lake +Apache Paimon — The core storage layer enabling ACID transactions, schema evolution, and unified batch-streaming reads +StarRocks & Apache Spark + Paimon — Delivering exceptional interactive and batch query performance directly on the lake +Apache Kyuubi — The unified, serverless SQL gateway tying it all together +Apache Gravitino — Unified metadata management, enabling centralized metadata governance across engines and data sources +Attendees will walk away with a proven methodology for building a unified Lakehouse around Paimon and other key components, along with practical guidance on component selection, integration, and production tuning across the ecosystem, as well as additional Data+AI application scenarios. + +### Speakers: + + +
+ +Zhoulong Liu: Senior Big Data Specialist, eclicktech + +The head of the Big Data Department at eclicktech previously worked at Sohu Video and Tencent and is skilled in building big data platforms and related systems. \ No newline at end of file diff --git a/content/sessions/datalake-1212404.zh.md b/content/sessions/datalake-1212404.zh.md new file mode 100644 index 000000000..8077129a0 --- /dev/null +++ b/content/sessions/datalake-1212404.zh.md @@ -0,0 +1,28 @@ +--- +title: "构建统一湖仓：Apache Paimon 及其生态的最佳实践" +date: "2026-08-07T14:30:00" +track: "datalake" +presenters: "Zhoulong Liu" +stype: "中文演讲" +room: "主会场 - 颐和厅" +--- + +随着数据架构不断演进，为批处理和流处理分别维护独立的孤岛——Lambda 架构的标志——已经变得越来越昂贵和复杂。我们如何才能构建一个真正统一的平台，既能提供实时的数据新鲜度，又能在规模化场景下实现高性能分析？ + +在本次演讲中，我们将超越理论，深入构建以 Apache Paimon 为中心的下一代统一湖仓的实战一线。我们将分享身经百战的最佳实践和真实落地模式，演示如何通过深度的生态协同来架构一条无缝的数据管道： + +Apache Flink + Paimon —— 稳健、低时延的实时入湖，进入事务型数据湖 +Apache Paimon —— 核心存储层，支持 ACID 事务、schema 演进和流批统一读取 +StarRocks 与 Apache Spark + Paimon —— 直接在湖上提供卓越的交互式和批处理查询性能 +Apache Kyuubi —— 把这一切串联起来的统一、无服务器 SQL 网关 +Apache Gravitino —— 统一元数据管理，支持跨引擎和数据源的集中化元数据治理 +听众将带走一套围绕 Paimon 及其他关键组件构建统一湖仓的、经过验证的方法论，以及跨生态的组件选型、集成和生产调优的实用指导，还有额外的 Data+AI 应用场景。 + +### 讲师: + + +
+ +Zhoulong Liu：eclicktech 资深大数据专家 + +作为 eclicktech 大数据部门的负责人，他此前曾在搜狐视频和腾讯工作，擅长构建大数据平台及相关系统。 \ No newline at end of file diff --git a/content/sessions/datalake-1212438.md b/content/sessions/datalake-1212438.md new file mode 100644 index 000000000..e83d51fd8 --- /dev/null +++ b/content/sessions/datalake-1212438.md @@ -0,0 +1,30 @@ +--- +title: "Challenges of Implementing Iceberg Features in a C++ Query Engine" +date: "2026-08-08T16:45:00" +track: "datalake" +presenters: "Zoltán Borók-Nagy, Péter Rózsa" +stype: "English Session" +room: "MainRoom - YiHe Hall" +--- + +Apache Impala is a hybrid, massively parallel query engine: its query planning and metadata handling are implemented in Java, while its execution engine is written in C++. This architecture presents unique challenges when integrating Apache Iceberg, whose reference implementation and ecosystem are primarily Java-based. Attendees will gain insight into the practical challenges of bringing Iceberg to non-JVM query engines. + +We will explore how Iceberg features are implemented in a distributed C++ query engine, and where Impala can reuse existing Iceberg libraries versus where custom implementations are required. We will dive into concrete examples, including metadata handling and row-level operations. The session will also highlight opportunities this model creates, such as rethinking how merge-on-read is implemented. + +Finally, attendees will also learn about the current state of Iceberg support in Impala and how our progress is going toward adopting Iceberg V3. + +### Speakers: + + +
+ +Zoltán Borók-Nagy: Cloudera, Principal Engineer + +Zoltán is a software engineer at Cloudera, working on Apache Impala. He is also a PMC member on the project. Currently he is leading the Impala/Iceberg integration efforts. Before Cloudera, Zoltán worked on C++ static analysis tools. He is interested in distributed, massively parallel systems, databases, and performance engineering. + + +
+ +Péter Rózsa: Cloudera, Software Engineer + +Péter Rózsa, a Software Engineer at Cloudera since 2020, started contributing to Apache Impala in 2022. He's been working on different parts of the Impala project, recently on Iceberg integrations \ No newline at end of file diff --git a/content/sessions/datalake-1212438.zh.md b/content/sessions/datalake-1212438.zh.md new file mode 100644 index 000000000..385c4730c --- /dev/null +++ b/content/sessions/datalake-1212438.zh.md @@ -0,0 +1,30 @@ +--- +title: "在 C++ 查询引擎中实现 Iceberg 特性的挑战" +date: "2026-08-08T16:45:00" +track: "datalake" +presenters: "Zoltán Borók-Nagy, Péter Rózsa" +stype: "英文演讲" +room: "主会场 - 颐和厅" +--- + +Apache Impala 是一个混合的大规模并行查询引擎：它的查询规划和元数据处理由 Java 实现，而执行引擎则用 C++ 编写。这种架构在集成 Apache Iceberg 时会带来独特的挑战——因为 Iceberg 的参考实现与生态主要以 Java 为基础。听众将了解到把 Iceberg 引入非 JVM 查询引擎时所面临的实际挑战。 + +我们将探讨 Iceberg 的特性如何在一个分布式 C++ 查询引擎中实现，以及 Impala 在哪些地方可以复用现有的 Iceberg 库、哪些地方又需要自定义实现。我们会深入具体案例，包括元数据处理和行级操作。本次演讲还会强调这种模式所带来的机遇，例如重新思考 merge-on-read 的实现方式。 + +最后，听众还将了解 Impala 对 Iceberg 支持的现状，以及我们朝着采用 Iceberg V3 所取得的进展。 + +### 讲师: + + +
+ +Zoltán Borók-Nagy：Cloudera，Principal Engineer + +Zoltán 是 Cloudera 的软件工程师，从事 Apache Impala 相关工作。他也是该项目的 PMC 成员。目前他负责牵头 Impala/Iceberg 的集成工作。在加入 Cloudera 之前，Zoltán 从事 C++ 静态分析工具的开发。他感兴趣的方向包括分布式系统、大规模并行系统、数据库以及性能工程。 + + +
+ +Péter Rózsa：Cloudera，软件工程师 + +Péter Rózsa 自 2020 年起在 Cloudera 担任软件工程师，于 2022 年开始为 Apache Impala 做贡献。他参与过 Impala 项目的多个部分，最近主要专注于 Iceberg 集成相关工作。 \ No newline at end of file diff --git a/content/sessions/datalake-1212598.md b/content/sessions/datalake-1212598.md new file mode 100644 index 000000000..8efb35f30 --- /dev/null +++ b/content/sessions/datalake-1212598.md @@ -0,0 +1,34 @@ +--- +title: "From Data Ingestion to Data Lake: Building a Modern Lakehouse with Apache SeaTunnel" +date: "2026-08-08T14:30:00" +track: "datalake" +presenters: "Lidong Dai" +stype: "Chinese Session" +room: "MainRoom - YiHe Hall" +--- + +Building a data lake is no longer just about choosing Iceberg, Hudi, or Paimon. +In real-world systems, the biggest challenge often lies one step earlier: how data reliably, efficiently, and continuously enters the lake. +In this session, we will explore how Apache SeaTunnel serves as a unified data ingestion and integration layer for modern data lake architectures. Starting from common pain points—multi-source data, Batch + CDC coexistence, schema evolution, and operational complexity—we will walk through how SeaTunnel simplifies data movement into data lakes and lakehouse systems. +Through real production scenarios, you will see how SeaTunnel connects transactional databases, message queues, and file systems into Iceberg- or lakehouse-based storage, enabling scalable, maintainable, and evolvable data platforms. The talk focuses on practical architecture decisions, not vendor-specific solutions. + +The session will cover: +- Typical data lake architecture evolution and common pitfalls +- The role of data integration in lake and lakehouse systems +- Apache SeaTunnel architecture and design principles +- End-to-end ingestion examples: databases, CDC, and streaming data into data lakes +- Operational considerations and best practices +- Roadmap of SeaTunnel in the data lake ecosystem + + + + + +### Speakers: + + +
+ +Lidong Dai: WhaleOps Technology co-founder + +Apache Incubator Mentor, Apache DolphinScheduler PMC member & Apache SeaTunnel PMC member \ No newline at end of file diff --git a/content/sessions/datalake-1212598.zh.md b/content/sessions/datalake-1212598.zh.md new file mode 100644 index 000000000..94dff0fe2 --- /dev/null +++ b/content/sessions/datalake-1212598.zh.md @@ -0,0 +1,33 @@ +--- +title: "从数据接入到数据湖：用 Apache SeaTunnel 构建现代湖仓" +date: "2026-08-08T14:30:00" +track: "datalake" +presenters: "Lidong Dai" +stype: "中文演讲" +room: "主会场 - 颐和厅" +--- + +构建数据湖，早已不再只是选择 Iceberg、Hudi 还是 Paimon 的问题。 +在真实系统中，最大的挑战往往更靠前一步：数据如何可靠、高效、持续地进入数据湖。 +在本次演讲中，我们将探讨 Apache SeaTunnel 如何作为现代数据湖架构的统一数据接入与集成层。从常见的痛点出发——多源数据、Batch 与 CDC 共存、schema 演进，以及运维复杂度——我们将走查 SeaTunnel 如何简化进入数据湖与湖仓系统的数据流转。 +通过真实的生产场景，你将看到 SeaTunnel 如何把事务型数据库、消息队列和文件系统接入基于 Iceberg 或湖仓的存储，从而构建可扩展、可维护、可演进的数据平台。本次演讲聚焦于务实的架构决策，而非特定厂商的方案。 + +本次演讲将涵盖： +- 典型的数据湖架构演进与常见陷阱 +- 数据集成在数据湖与湖仓系统中的角色 +- Apache SeaTunnel 的架构与设计原则 +- 端到端的接入示例：把数据库、CDC 和流式数据接入数据湖 +- 运维考量与最佳实践 +- SeaTunnel 在数据湖生态中的路线图 + + + + +### 讲师: + + +
+ +Lidong Dai：WhaleOps Technology 联合创始人 + +Apache 孵化器 Mentor，Apache DolphinScheduler PMC 成员 & Apache SeaTunnel PMC 成员 \ No newline at end of file diff --git a/content/sessions/datalake-1212641.md b/content/sessions/datalake-1212641.md new file mode 100644 index 000000000..b7d72be63 --- /dev/null +++ b/content/sessions/datalake-1212641.md @@ -0,0 +1,36 @@ +--- +title: "The Anatomy of Iceberg Failures: Lessons from Real-World Escalations" +date: "2026-08-07T16:45:00" +track: "datalake" +presenters: "Noémi Pap-Takács, Boglárka Egyed" +stype: "English Session" +room: "MainRoom - YiHe Hall" +--- + +Apache Iceberg has revolutionized data lakes by bringing ACID transactions and flexible table updates. However, at enterprise scale, "hands-off" management is a myth. Without the right maintenance strategy, high-throughput systems often suffer from rising storage costs, query performance degradation, and failures. + +In this session, we will dissect real-world customer escalations ranging from systems crashing because they had to process massive amounts of metadata, to cleanup tasks being blocked by constant data updates from multiple concurrent writers. + +Attendees will move beyond the documentation to learn a proven practical guide for switching from firefighting emergencies to proactive management, keeping the data lake healthy and performant. + +We will cover: +- Observability: Key metrics to monitor to detect health issues before they trigger an escalation, +- Tips for performance: Reducing storage and compute costs, +- Maintenance strategies: Configure maintenance features and schedule compaction jobs without locking out production writers, +- Recovery: How to repair tables and recover from failures. + +### Speakers: + + +
+ +Noémi Pap-Takács: Apache Impala Committer + +Noémi Pap-Takács is a software engineer at Cloudera and a committer on the Apache Impala project. Her focus lies in performance optimization and the integration of Apache Iceberg into Impala. + + +
+ +Boglárka Egyed: Engineering Director at Cloudera + +Boglárka Egyed is an Engineering Director at Cloudera, leading the teams behind Apache Impala and Hive. A former automotive engineer turned big data enthusiast, she spent years as an Apache Sqoop developer before moving into leadership. Today, she focuses on scaling engineering excellence and advancing open-source innovation for modern data architecture. \ No newline at end of file diff --git a/content/sessions/datalake-1212641.zh.md b/content/sessions/datalake-1212641.zh.md new file mode 100644 index 000000000..3eeb73c98 --- /dev/null +++ b/content/sessions/datalake-1212641.zh.md @@ -0,0 +1,36 @@ +--- +title: "Iceberg 故障剖析：来自真实线上升级事件的经验教训" +date: "2026-08-07T16:45:00" +track: "datalake" +presenters: "Noémi Pap-Takács, Boglárka Egyed" +stype: "英文演讲" +room: "主会场 - 颐和厅" +--- + +Apache Iceberg 通过引入 ACID 事务和灵活的表更新，彻底改变了数据湖。然而在企业级规模下，"放手不管"式的管理只是一种幻想。如果没有正确的维护策略，高吞吐系统往往会饱受存储成本上升、查询性能退化乃至故障的困扰。 + +在本次演讲中，我们将剖析真实的客户升级事件——从因不得不处理海量元数据而崩溃的系统，到因多个并发写入者持续更新数据而导致清理任务被阻塞的情况。 + +听众将超越文档，学到一份经过验证的实用指南：从四处"救火"转向主动管理，让数据湖保持健康和高性能。 + +我们将涵盖： +- 可观测性：需要监控的关键指标，以便在问题升级为事件之前发现健康隐患。 +- 性能技巧：降低存储与计算成本。 +- 维护策略：配置维护特性并调度 compaction 作业，同时不锁死生产写入者。 +- 恢复：如何修复表并从故障中恢复。 + +### 讲师: + + +
+ +Noémi Pap-Takács：Apache Impala Committer + +Noémi Pap-Takács 是 Cloudera 的软件工程师，也是 Apache Impala 项目的 committer。她的专长在于性能优化以及将 Apache Iceberg 集成到 Impala 中。 + + +
+ +Boglárka Egyed：Cloudera 工程总监 + +Boglárka Egyed 是 Cloudera 的工程总监，带领着 Apache Impala 和 Hive 背后的团队。她从汽车工程师转型为大数据爱好者，在转向管理岗位之前，她曾作为 Apache Sqoop 开发者工作多年。如今，她专注于扩大工程卓越性，并为现代数据架构推进开源创新。 \ No newline at end of file diff --git a/content/sessions/datalake-1213297.md b/content/sessions/datalake-1213297.md new file mode 100644 index 000000000..f8cd59109 --- /dev/null +++ b/content/sessions/datalake-1213297.md @@ -0,0 +1,23 @@ +--- +title: "Evolution to Lakehouse: Cost-Effective Hive Migration and Real-Time Ingestion at Scale" +date: "2026-08-08T15:00:00" +track: "datalake" +presenters: "Hangxiang Yu" +stype: "Chinese Session" +room: "MainRoom - YiHe Hall" +--- + +At Didi, scaling our Hive ecosystem to Iceberg required rebuilding our strategy of data integrity and governance beyond simple format conversion. We engineered a zero-downtime offline migration strategy using a snapshot-based approach, enabling consistency checks and instant rollbacks before cutover. To tackle metadata bloat and orphan files, we also integrated Apache Amoro for autonomous optimization. + +Building on this foundation, we extended our architecture to real-time scenarios. By streamlining the ingestion pipeline with Iceberg, we reduced data latency from hourly to minute-level. This unified approach not only simplified ETL pipelines but also saved petabytes of storage and significantly cut computing costs. + +This session offers actionable insights into building a stable, high-performance Lakehouse in production. + +### Speakers: + + +
+ +Hangxiang Yu: Apache Flink Committer & Real-Time Computing Team Lead at Didi + +Hangxiang Yu is an Apache Flink Committer and currently leads the real-time computing engineering team at Didi. With years of hands-on practice in distributed systems and storage, he manages Didi's real-time infrastructure, including Apache Flink, data ingestion pipelines, and P0-level real-time data warehousing. He is also focused on advancing Didi's Lakehouse architecture, working to improve data freshness, optimize infrastructure costs, and build self-managing governance capabilities. \ No newline at end of file diff --git a/content/sessions/datalake-1213297.zh.md b/content/sessions/datalake-1213297.zh.md new file mode 100644 index 000000000..7c229b3b1 --- /dev/null +++ b/content/sessions/datalake-1213297.zh.md @@ -0,0 +1,23 @@ +--- +title: "迈向湖仓：高性价比的 Hive 迁移与大规模实时入湖" +date: "2026-08-08T15:00:00" +track: "datalake" +presenters: "Hangxiang Yu" +stype: "中文演讲" +room: "主会场 - 颐和厅" +--- + +在滴滴，把我们的 Hive 生态扩展到 Iceberg，需要的不仅仅是简单的格式转换，而是重建我们的数据完整性与治理策略。我们设计了一套基于快照的零停机离线迁移策略，能够在切换之前进行一致性检查和即时回滚。为了应对元数据膨胀和孤儿文件问题，我们还集成了 Apache Amoro 来实现自主优化。 + +在这一基础上，我们把架构扩展到了实时场景。通过用 Iceberg 精简入湖管道，我们把数据延迟从小时级降低到了分钟级。这种统一的做法不仅简化了 ETL 管道，还节省了 PB 级的存储，并显著降低了计算成本。 + +本次演讲提供关于在生产环境中构建稳定、高性能湖仓的可落地洞见。 + +### 讲师: + + +
+ +Hangxiang Yu：Apache Flink Committer & 滴滴实时计算团队负责人 + +Hangxiang Yu 是 Apache Flink Committer，目前在滴滴带领实时计算工程团队。凭借在分布式系统和存储方面多年的实战经验，他管理着滴滴的实时基础设施，包括 Apache Flink、数据入湖管道以及 P0 级实时数据仓库。他还致力于推进滴滴的湖仓架构，努力提升数据新鲜度、优化基础设施成本，并构建自管理的治理能力。 \ No newline at end of file diff --git a/content/sessions/datalake-1213358.md b/content/sessions/datalake-1213358.md new file mode 100644 index 000000000..b1e24119a --- /dev/null +++ b/content/sessions/datalake-1213358.md @@ -0,0 +1,23 @@ +--- +title: "What's new in Apache Cloudberry (Incubating) 2.1 and 3.0" +date: "2026-08-07T15:45:00" +track: "datalake" +presenters: "Max Yang" +stype: "Chinese Session" +room: "MainRoom - YiHe Hall" +--- + +Apache Cloudberry is an advanced and mature open-source MPP database, derived from the open-source version of the Pivotal Greenplum Database® but built on a more modern PostgreSQL kernel and with more advanced enterprise capabilities. Cloudberry can serve as a data warehouse and can also be used for large-scale analytics and AI/ML workloads. +Since Apache Cloudberry entered the Apache Incubator and delivered its first incubating release in 2025, the project has continued to evolve rapidly. In 2026, the community is advancing both the 2.x and 3.x release lines, bringing significant enhancements across performance, architecture, and usability. +In this talk, we will provide an overview of the latest innovations in Apache Cloudberry. We will cover continued improvements in the 2.x series, built on PostgreSQL 14, including performance optimizations, new features, and ecosystem enhancements. We will also introduce the 3.x series, powered by PostgreSQL 16, and discuss the architectural opportunities enabled by the newer PostgreSQL kernel. +Topics will include selected improvements in query execution, storage, security, developer experience, and compatibility, along with a look at how these changes shape the future direction of Apache Cloudberry. +This session is intended for users, contributors, and anyone interested in the technical evolution of next-generation open-source MPP data warehouses. + +### Speakers: + + +
+ +Max Yang: Apache Cloudberry PPMC Member, Tech VP of HashData + +Apache Cloudberry PPMC Member, Tech VP of HashData \ No newline at end of file diff --git a/content/sessions/datalake-1213358.zh.md b/content/sessions/datalake-1213358.zh.md new file mode 100644 index 000000000..e3238de17 --- /dev/null +++ b/content/sessions/datalake-1213358.zh.md @@ -0,0 +1,23 @@ +--- +title: "Apache Cloudberry（孵化中）2.1 与 3.0 有哪些新特性" +date: "2026-08-07T15:45:00" +track: "datalake" +presenters: "Max Yang" +stype: "中文演讲" +room: "主会场 - 颐和厅" +--- + +Apache Cloudberry 是一款先进且成熟的开源 MPP 数据库，源自 Pivotal Greenplum Database® 的开源版本，但构建在更现代的 PostgreSQL 内核之上，并具备更高级的企业能力。Cloudberry 既可以用作数据仓库，也可用于大规模分析以及 AI/ML 工作负载。 +自 Apache Cloudberry 进入 Apache 孵化器并于 2025 年交付首个孵化版本以来，该项目持续快速演进。2026 年，社区同时推进 2.x 和 3.x 两条发布线，在性能、架构和易用性上带来了重大增强。 +在本次演讲中，我们将概述 Apache Cloudberry 的最新创新。我们将介绍基于 PostgreSQL 14 构建的 2.x 系列的持续改进，包括性能优化、新特性和生态增强。我们还会介绍基于 PostgreSQL 16 的 3.x 系列，并探讨更新的 PostgreSQL 内核所带来的架构机遇。 +话题将涵盖查询执行、存储、安全、开发者体验和兼容性方面的精选改进，并展望这些变化如何塑造 Apache Cloudberry 的未来方向。 +本次演讲面向用户、贡献者，以及任何对下一代开源 MPP 数据仓库技术演进感兴趣的人。 + +### 讲师: + + +
+ +Max Yang：Apache Cloudberry PPMC 成员，HashData 技术 VP + +Apache Cloudberry PPMC 成员，HashData 技术 VP \ No newline at end of file diff --git a/content/sessions/datalake-1213484.md b/content/sessions/datalake-1213484.md new file mode 100644 index 000000000..01cf39b60 --- /dev/null +++ b/content/sessions/datalake-1213484.md @@ -0,0 +1,27 @@ +--- +title: "Securing the Lakehouse: Identity-Aware Governance for Iceberg REST with Apache Gravitino" +date: "2026-08-09T15:15:00" +track: "datalake" +presenters: "Rory Qi" +stype: "English Session" +room: "MainRoom - YiHe Hall" +--- + +Apache Iceberg has become the de facto standard for open table formats, and the Iceberg REST Catalog is the key to decoupling compute from storage. However, while the REST specification defines how to exchange metadata, it leaves the security—authentication, authorization, and audit—as an exercise for the user. For organizations transitioning from experimental projects to production lakehouses, this "security gap" presents a significant hurdle. +In this talk, we introduce Apache REST, an open-source federated metadata lake that provides a production-ready implementation of the Iceberg REST service. We will walk through the technical architecture of how REST acts as a secure catalog proxy. +Key takeaways include: +Identity & Authentication: A deep dive into Gravitino’s support for OAuth2, Basic, Kerberos, and HTTPS to ensure only verified clients can access the catalog. +Unified Access Control: How to manage fine-grained permissions (RBAC) at the catalog, namespace, and table levels, even when using legacy backends like Hive Metastore or JDBC. +Credential Vending: How Gravitino securely provides short-lived storage credentials (S3/GCS/Azure) to clients, ensuring data-layer security is as strong as the metadata layer. +Real-world Integration: A demonstration of how Spark, Trino, and Flink can interact with a secured Gravitino Iceberg REST endpoint without custom engine forks or patches. +Whether you are a platform engineer building a multi-tenant data lake or a security architect concerned about open-source governance, this session will provide a blueprint for a secure, vendor-neutral Iceberg infrastructure + + +### Speakers: + + +
+ +Rory Qi: Apache Gravitino committer, Apache Uniffle PMC Chair + +Apache Gravitino PMC member, Apache Uniffle PMC chair, ASF member, Datastrato engineer, ever worked at Tencent, Baidu \ No newline at end of file diff --git a/content/sessions/datalake-1213484.zh.md b/content/sessions/datalake-1213484.zh.md new file mode 100644 index 000000000..b4751f79d --- /dev/null +++ b/content/sessions/datalake-1213484.zh.md @@ -0,0 +1,27 @@ +--- +title: "守护湖仓安全：用 Apache Gravitino 为 Iceberg REST 提供身份感知的治理" +date: "2026-08-09T15:15:00" +track: "datalake" +presenters: "Rory Qi" +stype: "英文演讲" +room: "主会场 - 颐和厅" +--- + +Apache Iceberg 已经成为开放表格式的事实标准，而 Iceberg REST Catalog 是实现存算分离的关键。然而，尽管 REST 规范定义了如何交换元数据，却把安全——认证、授权和审计——留给了用户自行解决。对于从试验性项目迈向生产级湖仓的组织而言，这一"安全缺口"是一个重大障碍。 +在本次演讲中，我们介绍 Apache REST——一个开源的联邦元数据湖，提供生产可用的 Iceberg REST 服务实现。我们将走查 REST 如何作为安全 catalog 代理的技术架构。 +核心收获包括： +身份与认证：深入剖析 Gravitino 对 OAuth2、Basic、Kerberos 和 HTTPS 的支持，确保只有经过验证的客户端才能访问 catalog。 +统一访问控制：如何在 catalog、namespace 和表级别管理细粒度权限（RBAC），即使在使用 Hive Metastore 或 JDBC 等传统后端时也能做到。 +凭据分发（Credential Vending）：Gravitino 如何安全地向客户端发放短时有效的存储凭据（S3/GCS/Azure），确保数据层的安全性与元数据层一样强。 +真实集成：演示 Spark、Trino 和 Flink 如何与受 Gravitino 安全保护的 Iceberg REST 端点交互，且无需对引擎进行自定义 fork 或打补丁。 +无论你是构建多租户数据湖的平台工程师，还是关注开源治理的安全架构师，本次演讲都将为构建一套安全、厂商中立的 Iceberg 基础设施提供蓝图。 + + +### 讲师: + + +
+ +Rory Qi：Apache Gravitino committer，Apache Uniffle PMC Chair + +Apache Gravitino PMC 成员，Apache Uniffle PMC Chair，ASF 成员，Datastrato 工程师，曾就职于腾讯、百度 \ No newline at end of file diff --git a/content/sessions/datalake-1214197.md b/content/sessions/datalake-1214197.md new file mode 100644 index 000000000..a2e268742 --- /dev/null +++ b/content/sessions/datalake-1214197.md @@ -0,0 +1,26 @@ +--- +title: "Beyond the Protocol: Productionizing the Iceberg REST Catalog at Enterprise Scale" +date: "2026-08-07T15:00:00" +track: "datalake" +presenters: "Jerry Shao" +stype: "Chinese Session" +room: "MainRoom - YiHe Hall" +--- + +While the Iceberg REST Catalog (IRC) provides a standardized protocol for table management, implementing it in a mission-critical enterprise environment reveals a significant "implementation gap." Moving from a legacy Hive Metastore (HMS) to a REST-based architecture requires more than just a new API—it demands a total rethink of security, performance, and data continuity. + +In this session, we share a field-tested solution for deploying IRC in production. We break down the transition into three critical pillars: + +Seamless Migration: How to move existing HMS-managed tables to IRC without business disruption or "stop-the-world" maintenance windows. +Modernizing Security: Strategies for transitioning from Hadoop-native Kerberos authentication and Ranger-based authorization to modern, cloud-native OAuth and centralized identity providers. +Performance at Scale: Shifting the commit mechanism from the client to the server opens new doors for optimization. We will explore server-side performance enhancements that reduce commit conflicts and improve query planning latency. +Attendees will walk away with a practical roadmap for closing the gap between the IRC specification and a secure, high-performance production deployment. + +### Speakers: + + +
+ +Jerry Shao: Datastrato, CTO + +Jerry Shao is the co-founder and CTO of Datastrato, focused on open source Big Data are for more than 10 years. He is an Apache member, committer and PMC member of Apache Spark and Apache Inlong, the original creator of Apache Gravitino. \ No newline at end of file diff --git a/content/sessions/datalake-1214197.zh.md b/content/sessions/datalake-1214197.zh.md new file mode 100644 index 000000000..b5ebb6f4b --- /dev/null +++ b/content/sessions/datalake-1214197.zh.md @@ -0,0 +1,26 @@ +--- +title: "超越协议：在企业级规模下让 Iceberg REST Catalog 走向生产" +date: "2026-08-07T15:00:00" +track: "datalake" +presenters: "Jerry Shao" +stype: "中文演讲" +room: "主会场 - 颐和厅" +--- + +尽管 Iceberg REST Catalog（IRC）为表管理提供了标准化的协议，但在关键任务的企业环境中落地它，会暴露出一个显著的"实现差距"。从传统的 Hive Metastore（HMS）迁移到基于 REST 的架构，需要的不仅仅是一个新的 API——它要求对安全、性能和数据连续性进行彻底的重新思考。 + +在本次演讲中，我们分享一套在生产环境中部署 IRC 的、经过实战检验的方案。我们把这次转型拆解为三大支柱： + +无缝迁移：如何在不造成业务中断、也不出现"stop-the-world"维护窗口的情况下，把现有由 HMS 管理的表迁移到 IRC。 +安全现代化：从 Hadoop 原生的 Kerberos 认证和基于 Ranger 的授权，迁移到现代的、云原生的 OAuth 和集中化身份提供者（identity provider）的策略。 +大规模性能：把提交（commit）机制从客户端转移到服务端，为优化打开了新的大门。我们将探讨服务端的性能增强，它们能减少提交冲突并改善查询规划延迟。 +听众将带走一份实用的路线图，用以弥合 IRC 规范与安全、高性能生产部署之间的差距。 + +### 讲师: + + +
+ +Jerry Shao：Datastrato CTO + +Jerry Shao 是 Datastrato 的联合创始人兼 CTO，专注于开源大数据领域已超过 10 年。他是 Apache 成员、Apache Spark 和 Apache Inlong 的 committer 和 PMC 成员，也是 Apache Gravitino 的原创建者。 \ No newline at end of file diff --git a/content/sessions/dataops-1207567.md b/content/sessions/dataops-1207567.md new file mode 100644 index 000000000..433b3efe8 --- /dev/null +++ b/content/sessions/dataops-1207567.md @@ -0,0 +1,29 @@ +--- +title: "From 'Usable' to 'Governable': ClassLoader Lifecycle Governance Practice in Apache SeaTunnel" +date: "2026-08-09T14:30:00" +track: "dataops" +presenters: "金祥 杨" +stype: "Chinese Session" +room: "Mtn WanShou Hall" +--- + +ClassLoader leaks are among the most hidden and difficult-to-diagnose runtime problems in long-lived JVM systems, and a common challenge faced by long-running JVM workloads across the Apache ecosystem. Existing approaches largely remain at the "post-mortem investigation" stage: monitoring tools and heap dumps can tell developers "which ClassLoaders are still alive," but struggle to clearly explain "why they cannot be reclaimed," let alone translate governance intent into verifiable, reproducible engineering practices. + +Through deep analysis of Apache SeaTunnel's classloading mechanism, we identified a widely overlooked blind spot: a large number of seemingly correct lifecycle implementations never establish explicit resource-close semantics. Without enforced close and drain constraints, resource reclamation becomes highly unpredictable — and in certain scenarios, leads to ClassLoaders that can never be collected. + +Drawing on our exploration of runtime governance for long-lived JVM systems, we proposed a systematic ClassLoader lifecycle governance improvement plan to the Apache SeaTunnel community. The core shift is from passive "post-hoc residual reference hunting" to proactive "building deterministic reclaimable semantics and lifecycle closure." We incrementally introduced explicit lifecycle close mechanisms, enforced classloading boundary constraints, and active residual reference cleanup. The governance proposal is currently under community discussion, with the Phase 1 optimization PR in review. + +This talk will explore real-world community practices in Apache SeaTunnel, covering: + +1. Why long-lived systems cannot rely on implicit GC to manage underlying runtime resources +2. How to build a ClassLoader governance standard for long-lived JVM systems and progressively land it in Apache open source projects +3. How to complete a smooth, kernel-level architectural governance upgrade without breaking compatibility + +### Speakers: + + +
+ +金祥 杨: Creator of LingFrame & LingMirror | Apache SeaTunnel Community Member | JVM Runtime Governance + +Jinxiang Yang is the creator of LingFrame (灵珑), an open source runtime governance framework for long-lived JVM single-process systems, and LingMirror, an IntelliJ IDEA plugin for static ClassLoader leak diagnosis. Focused on the governance of long-running JVM systems, he identified critical ClassLoader lifecycle gaps in Apache SeaTunnel through deep source code analysis, and proposed a systematic governance improvement plan that has been actively discussed in the SeaTunnel community. He believes a system is not complete when it runs — it is complete when its lifecycle can be observed, governed, and proven correct over time. \ No newline at end of file diff --git a/content/sessions/dataops-1207567.zh.md b/content/sessions/dataops-1207567.zh.md new file mode 100644 index 000000000..8f9327d85 --- /dev/null +++ b/content/sessions/dataops-1207567.zh.md @@ -0,0 +1,29 @@ +--- +title: "从'可用'到'可治'：Apache SeaTunnel 中的 ClassLoader 生命周期治理实践" +date: "2026-08-09T14:30:00" +track: "dataops" +presenters: "金祥 杨" +stype: "中文演讲" +room: "万寿山会议室" +--- + +在长生命周期的 JVM 系统中，ClassLoader 泄漏是最隐蔽、也最难诊断的一类运行时问题，同时也是整个 Apache 生态中长时运行 JVM 工作负载普遍面临的挑战。现有方案大多仍停留在"事后排查"阶段：监控工具和堆转储（heap dump）能告诉开发者"哪些 ClassLoader 还活着"，却很难讲清"为什么它们无法被回收"，更遑论把治理意图转化为可验证、可复现的工程实践。 + +通过对 Apache SeaTunnel 类加载机制的深入分析，我们发现了一个被广泛忽视的盲区：大量看似正确的生命周期实现，从未建立显式的资源关闭（close）语义。没有强制的 close 与 drain 约束，资源回收就变得高度不可预测——在某些场景下，甚至会导致 ClassLoader 永远无法被回收。 + +基于我们对长生命周期 JVM 系统运行时治理的探索，我们向 Apache SeaTunnel 社区提出了一套系统化的 ClassLoader 生命周期治理改进方案。其核心转变在于：从被动的"事后追查残留引用"，转向主动地"构建确定性的可回收语义与生命周期闭环"。我们渐进式地引入了显式的生命周期 close 机制、强制的类加载边界约束，以及主动的残留引用清理。该治理提案目前正在社区讨论中，第一阶段优化的 PR 处于评审阶段。 + +本次演讲将探讨 Apache SeaTunnel 中真实的社区实践，涵盖： + +1. 为什么长生命周期系统不能依赖隐式的 GC 来管理底层运行时资源 +2. 如何为长生命周期 JVM 系统构建 ClassLoader 治理标准，并逐步落地到 Apache 开源项目中 +3. 如何在不破坏兼容性的前提下，完成平滑的内核级架构治理升级 + +### 讲师: + + +
+ +金祥 杨：LingFrame 与 LingMirror 作者 | Apache SeaTunnel 社区成员 | JVM 运行时治理 + +Jinxiang Yang 是 LingFrame（灵珑）和 LingMirror 的作者。前者是面向长生命周期 JVM 单进程系统的开源运行时治理框架，后者是用于静态诊断 ClassLoader 泄漏的 IntelliJ IDEA 插件。他专注于长时运行 JVM 系统的治理，通过对源码的深入分析，定位到 Apache SeaTunnel 中关键的 ClassLoader 生命周期缺口，并提出了一套系统化的治理改进方案，目前在 SeaTunnel 社区被积极讨论。他相信：一个系统在"能跑"时并不算完成——只有当它的生命周期可被观测、可被治理，并能被证明长期正确时，才算真正完成。 \ No newline at end of file diff --git a/content/sessions/dataops-1210676.md b/content/sessions/dataops-1210676.md new file mode 100644 index 000000000..5025a5824 --- /dev/null +++ b/content/sessions/dataops-1210676.md @@ -0,0 +1,25 @@ +--- +title: "Why SeaTunnel Engine Needs Clearer Distributed Abstractions" +date: "2026-08-09T15:45:00" +track: "dataops" +presenters: "Doyeon Kim" +stype: "English Session" +room: "Mtn WanShou Hall" +--- + +In distributed data processing engines, responsibilities such as coordination, communication, and state management are often tightly coupled within a single framework. This can make the system difficult to evolve, extend, or replace individual components over time. + +In this talk, I will share my experience working on the SeaTunnel engine, where we explored introducing clearer abstractions to reduce dependency on a single distributed framework. I will discuss why this kind of decoupling is necessary, the approach we considered, and the challenges we encountered along the way. + +Rather than focusing on a specific technology choice, this talk will highlight the design questions and trade-offs involved in separating core responsibilities in a distributed engine. + +Attendees will gain practical insights into how to think about abstraction boundaries and how to approach architectural evolution in real-world distributed systems. + +### Speakers: + + +
+ +Doyeon Kim: Apache SeaTunnel Committer + +Doyeon Kim is an Apache SeaTunnel Committer and a student with a strong interest in data engineering and distributed systems. She has contributed to SeaTunnel in areas such as connector development, engine improvements, and architectural discussions. Her recent work has focused on engine internals, dependency decoupling, and design trade-offs in distributed systems. Through this talk, she shares lessons learned from exploring clearer abstractions in the SeaTunnel engine. \ No newline at end of file diff --git a/content/sessions/dataops-1210676.zh.md b/content/sessions/dataops-1210676.zh.md new file mode 100644 index 000000000..d29148af0 --- /dev/null +++ b/content/sessions/dataops-1210676.zh.md @@ -0,0 +1,25 @@ +--- +title: "为什么 SeaTunnel Engine 需要更清晰的分布式抽象" +date: "2026-08-09T15:45:00" +track: "dataops" +presenters: "Doyeon Kim" +stype: "英文演讲" +room: "万寿山会议室" +--- + +在分布式数据处理引擎中，协调、通信和状态管理等职责往往紧密耦合在单一框架之内。这会让系统随着时间推移变得难以演进、扩展，也难以替换其中的个别组件。 + +在本次演讲中，我将分享自己在 SeaTunnel 引擎上的工作经验——我们探索如何引入更清晰的抽象，以降低对单一分布式框架的依赖。我会讨论这种解耦为何必要、我们考虑过的方式，以及一路上遇到的挑战。 + +本次演讲不会聚焦于某项具体的技术选型，而是着重于在分布式引擎中分离核心职责时所涉及的设计问题与权衡。 + +听众将获得关于如何思考抽象边界、如何在真实分布式系统中推进架构演进的实用洞察。 + +### 讲师: + + +
+ +Doyeon Kim：Apache SeaTunnel Committer + +Doyeon Kim 是 Apache SeaTunnel Committer，也是一名学生，对数据工程与分布式系统抱有浓厚兴趣。她为 SeaTunnel 贡献过连接器开发、引擎改进以及架构讨论等方面的工作。她近期的工作聚焦于引擎内部实现、依赖解耦以及分布式系统中的设计权衡。在本次演讲中，她将分享在 SeaTunnel 引擎中探索更清晰抽象的经验教训。 \ No newline at end of file diff --git a/content/sessions/dataops-1212167.md b/content/sessions/dataops-1212167.md new file mode 100644 index 000000000..d0a6cc478 --- /dev/null +++ b/content/sessions/dataops-1212167.md @@ -0,0 +1,19 @@ +--- +title: "From Natural Language to Reliable Data Pipelines: Building an AI-Powered CLI for Apache SeaTunnel" +date: "2026-08-09T15:15:00" +track: "dataops" +presenters: "Xin Zhang" +stype: "Chinese Session" +room: "Mtn WanShou Hall" +--- + +Modern DataOps is not only about moving data faster, but also about making pipeline development easier, safer, and more accessible to engineers and data teams. In this talk, I will introduce seatunnel-cli, a new Python-based CLI for Apache SeaTunnel that generates HOCON pipeline configurations directly from natural language descriptions in both English and Chinese.The tool is designed as an AI-powered multi-agent workflow: Planner, Config Generator, Validator, and Auto-fix. It combines a three-tier knowledge base, a connector catalog automatically generated from SeaTunnel Java source code, dry-run validation, and iterative repair to help users produce runnable pipeline configurations with less manual effort. It also supports multiple LLM providers, persistent session memory, interactive exploration, and single-shot scripting usage.I will share the technical design behind the CLI, including how we extract and resolve connector metadata at scale, how validation and auto-fix loops improve configuration quality, and how this approach can reduce the barrier to building SeaTunnel jobs in real-world DataOps scenarios. The session will also cover practical lessons from testing across multilingual inputs, multiple model providers, and broken-config recovery cases. + +### Speakers: + + +
+ +Xin Zhang: Amaozn Web Services, Solution Architect + +Xin Zhang is an AWS Solutions Architect, responsible for solution consulting and design based on the AWS Cloud platform. He has a rich experience in R&D and architecture practice in the fields of system architecture, data warehousing, and real-time computing. \ No newline at end of file diff --git a/content/sessions/dataops-1212167.zh.md b/content/sessions/dataops-1212167.zh.md new file mode 100644 index 000000000..d975d945d --- /dev/null +++ b/content/sessions/dataops-1212167.zh.md @@ -0,0 +1,19 @@ +--- +title: "从自然语言到可靠的数据流水线：为 Apache SeaTunnel 打造 AI 驱动的 CLI" +date: "2026-08-09T15:15:00" +track: "dataops" +presenters: "Xin Zhang" +stype: "中文演讲" +room: "万寿山会议室" +--- + +现代 DataOps 不仅关乎更快地搬运数据，更关乎让流水线的开发对工程师和数据团队而言更简单、更安全、更易上手。在本次演讲中，我将介绍 seatunnel-cli——一款基于 Python、面向 Apache SeaTunnel 的新 CLI，它能够直接从中英文自然语言描述生成 HOCON 流水线配置。该工具被设计为一个 AI 驱动的多 Agent 工作流：包含 Planner、Config Generator、Validator 和 Auto-fix。它结合了三层知识库、从 SeaTunnel Java 源码自动生成的连接器目录、dry-run 校验以及迭代修复，帮助用户以更少的人工成本产出可运行的流水线配置。它还支持多种 LLM 提供方、持久的会话记忆、交互式探索，以及一次性脚本式用法。我将分享这款 CLI 背后的技术设计，包括我们如何大规模地抽取并解析连接器元数据、校验与自动修复循环如何提升配置质量，以及这种做法如何降低在真实 DataOps 场景中构建 SeaTunnel 作业的门槛。本次演讲还会涵盖在多语言输入、多种模型提供方以及损坏配置恢复等场景下测试所得的实践经验。 + +### 讲师: + + +
+ +Xin Zhang：Amazon Web Services 解决方案架构师 + +Xin Zhang 是 AWS 解决方案架构师，负责基于 AWS 云平台的解决方案咨询与设计。他在系统架构、数据仓库和实时计算领域拥有丰富的研发与架构实践经验。 \ No newline at end of file diff --git a/content/sessions/dataops-1212212.md b/content/sessions/dataops-1212212.md new file mode 100644 index 000000000..b842e74fd --- /dev/null +++ b/content/sessions/dataops-1212212.md @@ -0,0 +1,22 @@ +--- +title: "Workflow Orchestration in the AI Era: New Challenges and Practices for Data Processing at Scale" +date: "2026-08-09T16:15:00" +track: "dataops" +presenters: "LiWei Yang" +stype: "Chinese Session" +room: "Mtn WanShou Hall" +--- + +The rise of AI workloads—autonomous driving and large language models—is fundamentally reshaping data processing paradigms. Data has shifted from structured tables to petabyte-scale multimodal files; processing granularity has narrowed from partition-level to frame/document-level; and task dependency graphs have evolved from linear ETL pipelines into complex workflows with dynamic parallelism and conditional branching. Traditional scheduling systems, built around structured data, static DAGs, and homogeneous resources, are facing unprecedented challenges. + +This talk draws on Xiaomi's production experience of orchestrating millions of task instances daily across its autonomous driving and large language model businesses. We examine four core challenges—dynamic execution beyond static DAGs, heterogeneous resource orchestration across Spark/Flink/Script on YARN/K8s/GPU, million-scale scheduling throughput, and data-event-driven multi-condition triggering—and present our solutions. Key highlights include declarative + YAML-based workflow definitions, an architecture upgrade that delivered a 10x improvement in scheduling throughput, resource profiling and pool-based governance that boosted utilization by 50%, and our exploration of Agent-driven development and operations. + +### Speakers: + + +
+ +LiWei Yang: Software R&D Engineer + +Xiaomi Senior Software R&D Engineer, leading Data Engineering at the Big Data Development Platform. Contributor to Kestra, Apache Gravitino, and Apache DolphinScheduler. \ No newline at end of file diff --git a/content/sessions/dataops-1212212.zh.md b/content/sessions/dataops-1212212.zh.md new file mode 100644 index 000000000..e8ae8a9ec --- /dev/null +++ b/content/sessions/dataops-1212212.zh.md @@ -0,0 +1,21 @@ +--- +title: "AI 时代的工作流编排：规模化数据处理的全新挑战与实践" +date: "2026-08-09T16:15:00" +track: "dataops" +presenters: "LiWei Yang" +stype: "中文演讲" +room: "万寿山会议室" +--- + +AI 工作负载——自动驾驶与大语言模型——的兴起，正在从根本上重塑数据处理范式。数据已从结构化的表格转变为 PB 级的多模态文件；处理粒度从分区级细化到了帧级/文档级；任务依赖图也从线性的 ETL 流水线，演进为具备动态并行与条件分支的复杂工作流。传统调度系统围绕结构化数据、静态 DAG 与同构资源而建，正面临前所未有的挑战。 + +本次演讲基于小米在自动驾驶与大语言模型业务中、每天编排数百万任务实例的生产经验。我们审视四个核心挑战——超越静态 DAG 的动态执行、跨 YARN/K8s/GPU 上 Spark/Flink/脚本的异构资源编排、百万级调度吞吐，以及数据事件驱动的多条件触发——并给出我们的解决方案。亮点包括：声明式的基于 YAML 的工作流定义；一次带来调度吞吐 10 倍提升的架构升级；将利用率提升 50% 的资源画像与资源池化治理；以及我们在 Agent 驱动的开发与运维上的探索。 + +### 讲师: + + +
+ +LiWei Yang：软件研发工程师 + +小米高级软件研发工程师，负责大数据开发平台的数据工程方向。Kestra、Apache Gravitino 与 Apache DolphinScheduler 的贡献者。 \ No newline at end of file diff --git a/content/sessions/dataops-1214632.md b/content/sessions/dataops-1214632.md new file mode 100644 index 000000000..9964ef8eb --- /dev/null +++ b/content/sessions/dataops-1214632.md @@ -0,0 +1,29 @@ +--- +title: "Building AI's Data Artery: Architecture and Practices of Unified Multimodal Data Pipelines" +date: "2026-08-09T13:30:00" +track: "dataops" +presenters: "Xiaochen Zhou" +stype: "Chinese Session" +room: "Mtn WanShou Hall" +--- + +Abstract: +In the GenAI era, the massive flow of multimodal data demands a robust infrastructure, yet fragmented data pipelines have become a critical bottleneck for enterprises. At Tongcheng Travel, we historically operated 4 disjointed data pipeline services (Offline Sync, Real-time Lake Ingestion, legacy Sqoop, and a standalone SeaTunnel service). This fragmentation caused extremely high maintenance costs and hindered unified data governance. + +This session details how we successfully architected a unified "Data Artery" through platformization. We will explore how we consolidated the data entry points and built a true "Batch-Stream Unified" foundational architecture based on Apache SeaTunnel, comprehensively supporting data flows from traditional data warehouses to modern AI scenarios. + +Key Content: + +1. Breaking Data Silos: A deep dive into designing a unified multimodal data pipeline service powered by the Apache SeaTunnel engine, smoothly replacing and consolidating 4 legacy integration systems to achieve complete architectural standardization. +2. Compute Enhancement & AI Multimodal Empowerment: Exploring how to deeply integrate SeaTunnel's Transform mechanism with real-time stream processing capabilities to efficiently execute complex data cleaning and dynamic transformations. We will highlight hardcore support for AI workloads, including real-time parsing of unstructured data and Embedding preprocessing for LLMs. +3. Zero-Downtime Migration & Strict Validation: Sharing enterprise-grade practices on migrating massive legacy tasks. We will detail the "Dynamic Task Conversion and Bi-directional Data Reconciliation Mechanism" we designed to ensure zero data loss and a seamless transition for the business during the underlying architecture upgrade. +4. Future Cloud-Native Evolution: Looking ahead at the blueprint for multimodal unified data pipelines. We will discuss cloud-native containerized deployments on Kubernetes for elastic scaling, and how to deeply integrate with the LLM ecosystem to build a robust Data+AI foundation. + +### Speakers: + + +
+ +Xiaochen Zhou: Data Engineer @ Tongcheng Travel | Apache SeaTunnel Committer + +Xiaochen Zhou is a Data Engineer at Tongcheng Travel and an active Apache SeaTunnel Committer. In his current role, he specializes in designing, building, and optimizing high-performance data pipelines. Within the open-source community, he is deeply involved in the core development and technical evolution of Apache SeaTunnel. Recently, his focus has shifted to the intersection of Data and AI, where he is dedicated to architecting unified multimodal data pipelines for the GenAI era. \ No newline at end of file diff --git a/content/sessions/dataops-1214632.zh.md b/content/sessions/dataops-1214632.zh.md new file mode 100644 index 000000000..cb98d3e4e --- /dev/null +++ b/content/sessions/dataops-1214632.zh.md @@ -0,0 +1,29 @@ +--- +title: "打造 AI 的数据动脉：统一多模态数据流水线的架构与实践" +date: "2026-08-09T13:30:00" +track: "dataops" +presenters: "Xiaochen Zhou" +stype: "中文演讲" +room: "万寿山会议室" +--- + +摘要： +在 GenAI 时代，多模态数据的海量流动需要坚实的基础设施，然而割裂的数据流水线已成为企业的关键瓶颈。在同程旅行，我们历史上并行运行着 4 套互不相通的数据流水线服务（离线同步、实时入湖、老旧的 Sqoop，以及一套独立的 SeaTunnel 服务）。这种割裂带来了极高的维护成本，也阻碍了统一的数据治理。 + +本次演讲将详细讲解我们如何通过平台化，成功打造一条统一的"数据动脉"。我们将探讨如何整合数据入口，并基于 Apache SeaTunnel 构建真正"批流一体"的基础架构，全面支撑从传统数仓到现代 AI 场景的数据流转。 + +核心内容： + +1. 打破数据孤岛：深入讲解基于 Apache SeaTunnel 引擎设计的统一多模态数据流水线服务，平滑替换并整合 4 套老旧集成系统，实现彻底的架构标准化。 +2. 计算增强与 AI 多模态赋能：探讨如何将 SeaTunnel 的 Transform 机制与实时流处理能力深度融合，高效执行复杂的数据清洗与动态转换。我们将重点呈现对 AI 工作负载的硬核支撑，包括非结构化数据的实时解析，以及面向 LLM 的 Embedding 预处理。 +3. 零停机迁移与严格校验：分享海量存量任务迁移的企业级实践。我们将详细讲解所设计的"动态任务转换与双向数据比对（reconciliation）机制"，确保在底层架构升级过程中零数据丢失，并实现业务的无缝过渡。 +4. 面向未来的云原生演进：展望多模态统一数据流水线的蓝图。我们将讨论基于 Kubernetes 的云原生容器化部署以实现弹性伸缩，以及如何与 LLM 生态深度融合，构建坚实的 Data+AI 底座。 + +### 讲师: + + +
+ +Xiaochen Zhou：同程旅行数据工程师 | Apache SeaTunnel Committer + +Xiaochen Zhou 是同程旅行的数据工程师，也是活跃的 Apache SeaTunnel Committer。在当前岗位上，他专注于设计、构建和优化高性能数据流水线。在开源社区中，他深度参与 Apache SeaTunnel 的核心开发与技术演进。近来，他的重心转向数据与 AI 的交汇领域，致力于为 GenAI 时代架构统一的多模态数据流水线。 \ No newline at end of file diff --git a/content/sessions/dataops-1214716.md b/content/sessions/dataops-1214716.md new file mode 100644 index 000000000..2b79504cf --- /dev/null +++ b/content/sessions/dataops-1214716.md @@ -0,0 +1,19 @@ +--- +title: "Enhancing Metadata Discovery: Gravitino + OpenSearch" +date: "2026-08-09T14:00:00" +track: "dataops" +presenters: "Yu Hui" +stype: "Chinese Session" +room: "Mtn WanShou Hall" +--- + +Gravitino provides a unified metadata layer across different data systems . In this talk, we show how to use OpenSearch to index and search metadata from Gravitino, enabling fast and simple metadata discovery. + +### Speakers: + + +
+ +Yu Hui: Software Engineer, Datastrato (Shanghai) + +Software Engineer at Datastrato (Shanghai), working on data infrastructure and metadata systems. Contributor to Apache Gravitino, focusing on unified catalog, metadata federation, and integration with query engines like Trino. \ No newline at end of file diff --git a/content/sessions/dataops-1214716.zh.md b/content/sessions/dataops-1214716.zh.md new file mode 100644 index 000000000..9550a4eb7 --- /dev/null +++ b/content/sessions/dataops-1214716.zh.md @@ -0,0 +1,19 @@ +--- +title: "增强元数据发现能力：Gravitino + OpenSearch" +date: "2026-08-09T14:00:00" +track: "dataops" +presenters: "Yu Hui" +stype: "中文演讲" +room: "万寿山会议室" +--- + +Gravitino 在不同的数据系统之上提供统一的元数据层。在本次演讲中，我们将展示如何使用 OpenSearch 对来自 Gravitino 的元数据进行索引与检索，从而实现快速、简便的元数据发现。 + +### 讲师: + + +
+ +Yu Hui：软件工程师，Datastrato（上海） + +Datastrato（上海）软件工程师，从事数据基础设施与元数据系统的工作。Apache Gravitino 贡献者，专注于统一目录、元数据联邦，以及与 Trino 等查询引擎的集成。 \ No newline at end of file diff --git a/content/sessions/datastorage-1186318.md b/content/sessions/datastorage-1186318.md new file mode 100644 index 000000000..a848b48de --- /dev/null +++ b/content/sessions/datastorage-1186318.md @@ -0,0 +1,28 @@ +--- +title: "Unleashing Native Performance: The Architecture and Roadmap of Apache Iceberg-cpp" +date: "2026-08-08T16:15:00" +track: "datastorage" +presenters: "Gang Wu, Junwang Zhao" +stype: "Chinese Session" +room: "Mtn WanShou Hall" +--- + +As modern data engines shift toward native execution (C++, Rust, and SIMD-optimized processing), the need for a high-performance, JVM-independent Iceberg implementation has never been greater. This session dives into the origins and design philosophy of iceberg-cpp, the official C++ implementation of the Apache Iceberg specification. + +We will share why Singdata initiated this project—targeting high-performance Iceberg implementation and seamless integration with native query engines. We will cover the project’s current progress, including REST Catalog support and Arrow-native integration, while outlining the roadmap for write support and V3 specification compliance. Furthermore, Singdata will share its internal journey of integrating iceberg-cpp into its high-performance stack and our commitment to long-term maintenance. This session is a call to action for C++ engine developers and performance enthusiasts to join us in building the fastest way to interact with the Iceberg ecosystem. + +### Speakers: + + +
+ +Gang Wu: Software Engineer at Singdata + +Gang is a software engineer at Singdata working on the lakehouse storage. He is a PMC member of Apache Arrow, Apache ORC, Apache Parquet, and a committer of Apache Iceberg. He is also an ASF member. Prior to Singdata, he has worked at Uber on Apache Spark and Alibaba on the MaxCompute cloud data warehouse, respectively. + + +
+ +Junwang Zhao: Database Engineer at Ant Group + +Junwang is a database engineer at Ant Group and contributes to PostgreSQL development in his spare time, where he is recognized as a PostgreSQL Significant Contributor. He is also a co-creator of the iceberg-cpp project, focusing on building open, high-performance data infrastructure. \ No newline at end of file diff --git a/content/sessions/datastorage-1186318.zh.md b/content/sessions/datastorage-1186318.zh.md new file mode 100644 index 000000000..97e3555a2 --- /dev/null +++ b/content/sessions/datastorage-1186318.zh.md @@ -0,0 +1,28 @@ +--- +title: "释放原生性能：Apache Iceberg-cpp 的架构与路线图" +date: "2026-08-08T16:15:00" +track: "datastorage" +presenters: "Gang Wu, Junwang Zhao" +stype: "中文演讲" +room: "万寿山会议室" +--- + +随着现代数据引擎转向原生执行（C++、Rust 以及 SIMD 优化的处理），业界对高性能、不依赖 JVM 的 Iceberg 实现的需求达到了前所未有的高度。本次演讲将深入剖析 iceberg-cpp 的起源与设计理念——这是 Apache Iceberg 规范的官方 C++ 实现。 + +我们将分享 Singdata 发起该项目的初衷——瞄准高性能的 Iceberg 实现以及与原生查询引擎的无缝集成。我们会介绍项目当前的进展，包括对 REST Catalog 的支持和 Arrow 原生集成，并勾勒写入支持与 V3 规范合规的路线图。此外，Singdata 还将分享把 iceberg-cpp 集成进其高性能技术栈的内部历程，以及我们对长期维护的承诺。本次演讲也是向 C++ 引擎开发者和性能爱好者发出的一封邀请函——欢迎加入我们，共同打造与 Iceberg 生态交互的最快方式。 + +### 讲师: + + +
+ +Gang Wu：Singdata 软件工程师 + +Gang 是 Singdata 的软件工程师，从事湖仓存储相关工作。他是 Apache Arrow、Apache ORC、Apache Parquet 的 PMC 成员，也是 Apache Iceberg 的 Committer。他同时是 ASF 成员。加入 Singdata 之前，他先后在 Uber 从事 Apache Spark、在阿里巴巴从事 MaxCompute 云数据仓库的工作。 + + +
+ +Junwang Zhao：蚂蚁集团数据库工程师 + +Junwang 是蚂蚁集团的数据库工程师，业余时间参与 PostgreSQL 的开发，并被授予 PostgreSQL Significant Contributor 称号。他也是 iceberg-cpp 项目的共同创建者之一，致力于构建开放、高性能的数据基础设施。 \ No newline at end of file diff --git a/content/sessions/datastorage-1199721.md b/content/sessions/datastorage-1199721.md new file mode 100644 index 000000000..b95e61ab2 --- /dev/null +++ b/content/sessions/datastorage-1199721.md @@ -0,0 +1,19 @@ +--- +title: "Apache Auron (Incubating): Working Progress and Community Development" +date: "2026-08-08T17:15:00" +track: "datastorage" +presenters: "Zhang Li" +stype: "Chinese Session" +room: "Mtn WanShou Hall" +--- + +This session will cover the latest development progress of Apache Auron (Incubating) project, focusing on two key advancements: 1. Auron on GPU - optimizing performance through GPU acceleration, and 2. Flink on Auron - integration with Apache Flink for enhanced stream processing capabilities. Additionally, I will share insights on community development, strategies and how to contribute to this incubating project. + +### Speakers: + + +
+ +Zhang Li: PPMC member of Apache Auron (Incubating) project + +Zhang Li, from the data platform team at Kuaishou Inc., is a PPMC member and lead developer of the Apache Auron (incubating) project. With 10 years of experience in big data architecture development, he currently focuses on enhancing the performance of Apache Spark and Auron. \ No newline at end of file diff --git a/content/sessions/datastorage-1199721.zh.md b/content/sessions/datastorage-1199721.zh.md new file mode 100644 index 000000000..fd1d60fc7 --- /dev/null +++ b/content/sessions/datastorage-1199721.zh.md @@ -0,0 +1,19 @@ +--- +title: "Apache Auron（孵化中）：研发进展与社区发展" +date: "2026-08-08T17:15:00" +track: "datastorage" +presenters: "Zhang Li" +stype: "中文演讲" +room: "万寿山会议室" +--- + +本次演讲将介绍 Apache Auron（孵化中）项目的最新研发进展，重点聚焦两项关键进展：1. Auron on GPU——通过 GPU 加速优化性能；2. Flink on Auron——与 Apache Flink 集成以增强流处理能力。此外，我还会分享关于社区发展的思考、策略，以及如何参与到这个孵化项目中来。 + +### 讲师: + + +
+ +Zhang Li：Apache Auron（孵化中）项目 PPMC 成员 + +Zhang Li 来自快手数据平台团队，是 Apache Auron（孵化中）项目的 PPMC 成员与主要开发者。他拥有 10 年大数据架构开发经验，目前专注于提升 Apache Spark 与 Auron 的性能。 \ No newline at end of file diff --git a/content/sessions/datastorage-1202774.md b/content/sessions/datastorage-1202774.md new file mode 100644 index 000000000..53e6b2b22 --- /dev/null +++ b/content/sessions/datastorage-1202774.md @@ -0,0 +1,27 @@ +--- +title: "How Celeborn Makes Shuffle in AlibabaCloud Serverless Spark Comfortable, Reassuring, and Worry-Free" +date: "2026-08-07T16:15:00" +track: "datastorage" +presenters: "Zhentao Shuai, Jiashu Xiong" +stype: "Chinese Session" +room: "Mtn WanShou Hall" +--- + +How AlibabCloud EMR Serverless Spark integrates Apache Celeborn to overcome traditional Spark Shuffle limitations (coupled storage/compute, connection storms, random I/O, and single-point failures), delivering a robust production experience. + +### Speakers: + + +
+ +Zhentao Shuai: Alibaba Cloud + +Software Development Engineer at Alibaba Cloud +Celeborn committer + + +
+ +Jiashu Xiong: Alibaba Cloud, Senior Development Engineer + +Apache Celeborn PMC member, mainly focused on the optimization of Apache Celeborn and the integration of Apache Celeborn with engines such as Flink and Spark \ No newline at end of file diff --git a/content/sessions/datastorage-1202774.zh.md b/content/sessions/datastorage-1202774.zh.md new file mode 100644 index 000000000..85cc746ca --- /dev/null +++ b/content/sessions/datastorage-1202774.zh.md @@ -0,0 +1,27 @@ +--- +title: "Celeborn 如何让阿里云 Serverless Spark 的 Shuffle 省心、放心、无忧" +date: "2026-08-07T16:15:00" +track: "datastorage" +presenters: "Zhentao Shuai, Jiashu Xiong" +stype: "中文演讲" +room: "万寿山会议室" +--- + +介绍阿里云 EMR Serverless Spark 如何通过集成 Apache Celeborn，克服传统 Spark Shuffle 的种种局限（存算耦合、连接风暴、随机 I/O 以及单点故障），从而带来稳健的生产体验。 + +### 讲师: + + +
+ +Zhentao Shuai：阿里云 + +阿里云软件开发工程师 +Celeborn Committer + + +
+ +Jiashu Xiong：阿里云，高级开发工程师 + +Apache Celeborn PMC 成员，主要专注于 Apache Celeborn 的优化，以及 Apache Celeborn 与 Flink、Spark 等引擎的集成。 \ No newline at end of file diff --git a/content/sessions/datastorage-1209471.md b/content/sessions/datastorage-1209471.md new file mode 100644 index 000000000..db67dd8d9 --- /dev/null +++ b/content/sessions/datastorage-1209471.md @@ -0,0 +1,35 @@ +--- +title: "Optimizing Apache YARN Clusters Using Intelligent Overcommitment" +date: "2026-08-07T16:45:00" +track: "datastorage" +presenters: "Sumit Puri, Wasif Khan" +stype: "English Session" +room: "Mtn WanShou Hall" +--- + +Apache YARN clusters running compute frameworks such as Apache Spark, Apache Tez, and MapReduce often face major inefficiencies due to static, user-defined resource allocations. This leads to over-provisioning, low utilization, and increased infrastructure costs, while requiring continuous manual tuning to maintain performance. +This talk introduces a method to Optimize Yarn, which improves cluster efficiency through a real-time, intelligent approach to resource management. By understanding workload behavior over time and combining it with live cluster conditions, the solution can make dynamic decisions that improve how resources are utilized across nodes without requiring application changes. +The solution is designed with planned safeguards to ensure stability and reliability in production environments, enabling organizations to adopt it without risking application performance or service-level objectives. +This talk shares a result-driven approach that delivered substantial improvements. Container throughput increased from under by 100%, while application throughput improved from approximately 28.5%. At the same time, the number of active YARN nodes required was reduced by 33%, directly resulting in a reduction of the infrastructure footprint. These gains were driven by significantly improved CPU and Memory utilization across the cluster. +By enabling smarter, real-time resource allocation, this methodology can help organizations increase throughput, reduce infrastructure costs, and simplify cluster operations. + +Refer to this Blog for a detailed explanation along with a video demo : https://engineering.acceldata.io/how-we-optimized-yarn-clusters-using-intelligent-overcommitment/ + + +### Speakers: + + +
+ +Sumit Puri: Senior Software Engineer, Acceldata Inc. + +Sumit Puri is a Senior Software Engineer at Acceldata, working on large-scale data infrastructure and resource optimization systems. His work focuses on improving cluster efficiency across YARN and Kubernetes environments through intelligent resource management, real-time telemetry, and predictive modeling. + +He has hands-on experience building distributed systems using Go, Kubernetes, and big data frameworks such as Spark and Tez. His recent work explores node-level optimization strategies, workload fingerprinting, and adaptive overcommitment techniques to maximize infrastructure utilization while maintaining stability. + + +
+ +Wasif Khan: Principal Engineer, Acceldata + +Principal Engineer at Acceldata, working on big data observability solutions \ No newline at end of file diff --git a/content/sessions/datastorage-1209471.zh.md b/content/sessions/datastorage-1209471.zh.md new file mode 100644 index 000000000..962d7c0b7 --- /dev/null +++ b/content/sessions/datastorage-1209471.zh.md @@ -0,0 +1,35 @@ +--- +title: "用智能超分（Overcommitment）优化 Apache YARN 集群" +date: "2026-08-07T16:45:00" +track: "datastorage" +presenters: "Sumit Puri, Wasif Khan" +stype: "英文演讲" +room: "万寿山会议室" +--- + +运行 Apache Spark、Apache Tez 和 MapReduce 等计算框架的 Apache YARN 集群，常因静态的、由用户定义的资源分配而出现严重的低效。这会导致过度配置、利用率低下和基础设施成本攀升，同时需要持续的人工调优才能维持性能。 +本次演讲介绍一种优化 YARN 的方法，通过实时、智能的资源管理来提升集群效率。该方法通过理解工作负载随时间变化的行为，并结合集群的实时状况做出动态决策，无需改动应用即可改善资源在各节点间的利用方式。 +该方案在设计上内置了周全的防护措施，以确保在生产环境中的稳定性与可靠性，让组织能够在不损害应用性能或服务级别目标（SLO）的前提下采纳它。 +本次演讲分享了一套以结果为导向、带来显著成效的方法：容器吞吐量提升了 100%，应用吞吐量提升了约 28.5%。与此同时，所需的活跃 YARN 节点数减少了 33%，直接缩减了基础设施占用。这些收益源于全集群范围内 CPU 与内存利用率的显著改善。 +通过实现更智能、实时的资源分配，这套方法可以帮助组织提升吞吐量、降低基础设施成本，并简化集群运维。 + +详细讲解及视频 Demo 请参见此博客：https://engineering.acceldata.io/how-we-optimized-yarn-clusters-using-intelligent-overcommitment/ + + +### 讲师: + + +
+ +Sumit Puri：Acceldata Inc. 高级软件工程师 + +Sumit Puri 是 Acceldata 的高级软件工程师，从事大规模数据基础设施与资源优化系统的工作。他的工作重心是通过智能资源管理、实时遥测与预测建模，提升 YARN 与 Kubernetes 环境下的集群效率。 + +他在使用 Go、Kubernetes，以及 Spark、Tez 等大数据框架构建分布式系统方面有丰富的一手经验。他近期的工作探索了节点级优化策略、工作负载指纹识别以及自适应超分技术，以在保持稳定性的同时最大化基础设施利用率。 + + +
+ +Wasif Khan：Acceldata 首席工程师 + +Acceldata 首席工程师，从事大数据可观测性解决方案的工作。 \ No newline at end of file diff --git a/content/sessions/datastorage-1210477.md b/content/sessions/datastorage-1210477.md new file mode 100644 index 000000000..d11ded463 --- /dev/null +++ b/content/sessions/datastorage-1210477.md @@ -0,0 +1,26 @@ +--- +title: "Apache Ozone: Auto File Expiration and Removing through S3 Lifecycle Configuration" +date: "2026-08-07T14:00:00" +track: "datastorage" +presenters: "Chen Xi, Sammi Chen" +stype: "Chinese Session" +room: "Mtn WanShou Hall" +--- + +Apache Ozone is a distributed storage system in the Hadoop ecosystem. It supports both Hadoop Filesystem and AWS S3 compatible RESTFull API. In a distributed storage system in the big data ecosystem, normally some files are short lived temporary data, and some files are long lived permanent data, system administrators or users have to manually remove the temporary files when they are not used anymore. In this session, we will introduce the new feature S3 Lifecycle Configuration - Object Expiration, which will be released soon in the next Ozone 2.2.0 release, about how the S3 Lifecycle Configuration - Object Expiration feature is designed, how to use the feature to automatically expiration and removing temporary files by setting lifecycle configuration on target bucket, and share the experience of this feature in production environment. + +### Speakers: + + +
+ +Chen Xi: Storage Expert Engineer, Data Infrastructure at Shopee + +Experienced in the storage domain, with a strong focus on distributed storage systems and performance optimization. Currently working in Shopee’s Data Infrastructure team, responsible for storage product development. + + +
+ +Sammi Chen: Cloudera Principal Storage Engineer + +Cloudera principal storage engineer, focusing on Apache Hadoop and Apache Ozone kernel development, currently being the Chair of Ozone PMC and Hadoop PMC, former big data storage tech leader of Tencent and Intel. \ No newline at end of file diff --git a/content/sessions/datastorage-1210477.zh.md b/content/sessions/datastorage-1210477.zh.md new file mode 100644 index 000000000..2e014d7eb --- /dev/null +++ b/content/sessions/datastorage-1210477.zh.md @@ -0,0 +1,26 @@ +--- +title: "Apache Ozone：基于 S3 生命周期配置的文件自动过期与删除" +date: "2026-08-07T14:00:00" +track: "datastorage" +presenters: "Chen Xi, Sammi Chen" +stype: "中文演讲" +room: "万寿山会议室" +--- + +Apache Ozone 是 Hadoop 生态中的分布式存储系统，同时支持 Hadoop 文件系统 API 和与 AWS S3 兼容的 RESTful API。在大数据生态的分布式存储系统中，通常有些文件是短生命周期的临时数据，有些则是长生命周期的持久化数据，系统管理员或用户不得不在临时文件不再使用时手动将其删除。在本次演讲中，我们将介绍即将在下一个 Ozone 2.2.0 版本中发布的新特性 S3 Lifecycle Configuration - 对象过期（Object Expiration），包括该特性的设计思路、如何通过对目标 bucket 设置生命周期配置来自动过期并删除临时文件，以及在生产环境中使用该特性的经验分享。 + +### 讲师: + + +
+ +Chen Xi：存储专家工程师，Shopee 数据基础设施团队 + +深耕存储领域，专注于分布式存储系统与性能优化。目前在 Shopee 数据基础设施团队工作，负责存储产品研发。 + + +
+ +Sammi Chen：Cloudera 首席存储工程师 + +Cloudera 首席存储工程师，专注于 Apache Hadoop 与 Apache Ozone 内核开发，现任 Ozone PMC 与 Hadoop PMC 主席，曾任腾讯与 Intel 的大数据存储技术负责人。 \ No newline at end of file diff --git a/content/sessions/datastorage-1211330.md b/content/sessions/datastorage-1211330.md new file mode 100644 index 000000000..3d020976d --- /dev/null +++ b/content/sessions/datastorage-1211330.md @@ -0,0 +1,23 @@ +--- +title: "The Consistency Model in Apache Ratis" +date: "2026-08-08T14:00:00" +track: "datastorage" +presenters: "Tsz-Wo Nicholas Sze" +stype: "English Session" +room: "Mtn WanShou Hall" +--- + +Apache Ratis is an open source Java library for the Raft Consensus Protocol, a distributed consensus algorithm ensuring that multiple machines work together as a single, coherent unit. Ratis constitutes a high performance implementation of Raft. To accommodate a diverse array of use cases, Ratis supports various configurable consistency levels. Furthermore, Ratis offers both synchronous and asynchronous APIs for read and write operations. Numerous projects, such as Apache Ozone, Apache IoTDB, Apache Celeborn, and Alluxio, utilize Ratis. + +Raft along with Ratis are designed to provide strong consistency; specifically, once a write operation is acknowledged as complete, all subsequent reads are guaranteed to reflect the effects of that write. Read operations are more interesting since there is a fundamental trade-off between consistency and performance. Ratis has implemented several read mechanisms, including direct read from the leader, linearizable read, read from the followers, read-after-write consistency and stale-read from followers. + +This presentation will commence with a concise introduction to the Raft protocol. Subsequently, we will examine the consistency model within Ratis and detail the associated APIs and configurations. + +### Speakers: + + +
+ +Tsz-Wo Nicholas Sze: Principal Software Engineer + +Dr. Tsz-Wo Nicholas Sze serves as a Principal Engineer at Cloudera. He holds the position of PMC Chair for Apache Ratis and is a PMC member for Apache Hadoop and Apache Ozone. His involvement with Apache Hadoop and HDFS dates back to 2007, making him one of the few initial contributors still active in the project. In 2016, he initiated the Apache Ratis project to facilitate high availability and high-performance write pipelines within Apache Ozone. Apache Ratis has since become one of the most widely adopted Raft Java libraries. He earned his Ph.D. degree in Computer Science from the University of Maryland College Park in 2007. \ No newline at end of file diff --git a/content/sessions/datastorage-1211330.zh.md b/content/sessions/datastorage-1211330.zh.md new file mode 100644 index 000000000..7f517b925 --- /dev/null +++ b/content/sessions/datastorage-1211330.zh.md @@ -0,0 +1,23 @@ +--- +title: "Apache Ratis 中的一致性模型" +date: "2026-08-08T14:00:00" +track: "datastorage" +presenters: "Tsz-Wo Nicholas Sze" +stype: "英文演讲" +room: "万寿山会议室" +--- + +Apache Ratis 是一个实现 Raft 共识协议的开源 Java 库，Raft 是一种分布式共识算法，用于确保多台机器像一个协调一致的整体那样协同工作。Ratis 是 Raft 的高性能实现。为适应多种多样的使用场景，Ratis 支持多种可配置的一致性级别。此外，Ratis 为读写操作同时提供了同步与异步 API。众多项目都使用了 Ratis，例如 Apache Ozone、Apache IoTDB、Apache Celeborn 和 Alluxio。 + +Raft 与 Ratis 的设计目标都是提供强一致性——具体而言，一旦某次写操作被确认完成，此后所有的读操作都保证能反映这次写操作的效果。读操作则更有意思，因为一致性与性能之间存在根本性的权衡。Ratis 实现了多种读机制，包括从 leader 直接读取、线性一致读（linearizable read）、从 follower 读取、读后写（read-after-write）一致性以及从 follower 读取陈旧数据（stale-read）。 + +本次演讲将先简要介绍 Raft 协议，随后考察 Ratis 中的一致性模型，并详细说明相关的 API 与配置。 + +### 讲师: + + +
+ +Tsz-Wo Nicholas Sze：首席软件工程师 + +Tsz-Wo Nicholas Sze 博士在 Cloudera 担任首席工程师。他是 Apache Ratis 的 PMC 主席，也是 Apache Hadoop 与 Apache Ozone 的 PMC 成员。他参与 Apache Hadoop 和 HDFS 的工作可追溯到 2007 年，是该项目为数不多、至今仍活跃的早期贡献者之一。2016 年，他发起了 Apache Ratis 项目，以支撑 Apache Ozone 中的高可用与高性能写入流水线。如今，Apache Ratis 已成为使用最广泛的 Raft Java 库之一。他于 2007 年获得马里兰大学帕克分校计算机科学博士学位。 \ No newline at end of file diff --git a/content/sessions/datastorage-1211771.md b/content/sessions/datastorage-1211771.md new file mode 100644 index 000000000..ffc346be5 --- /dev/null +++ b/content/sessions/datastorage-1211771.md @@ -0,0 +1,81 @@ +--- +title: "HBase Function Iteration and Architecture Evolution Practice at Xiaomi" +date: "2026-08-07T15:45:00" +track: "datastorage" +presenters: "Yuting sun, Yiming Gong" +stype: "Chinese Session" +room: "Mtn WanShou Hall" +--- + +2. Description： +This session will introduce the function iteration and deployment architecture evolution of HBase system at Xiaomi Group, as well as the work done in stability, performance and cost optimization, including dictionary compression, data structuring, SQL engine capabilities, containerization, and storage-compute separation. +3. Presentation Outline +3.1 Data Compression and Storage Optimization +Dictionary Compression Technology +- Core Principle: Build auxiliary compression structure based on high-frequency byte sequences +- Technical Features: Compression/decompression using the same dictionary, particularly suitable for structured data with repetitive patterns +- Practical Results: Significantly reduce storage costs and improve I/O efficiency +3.2 Observability and Governance Capabilities +Table Lineage Feature +- Functionality: Implement lightweight table lineage collection capability +- Collected Information: Client version, IP address, Kerberos account, and other key metadata +- Application Value: Support one-click lineage query for data governance and issue troubleshooting +3.3 Consistency and Availability Trade-offs +HBase AP Practice Exploration +- Positioning Shift: Exploration from CP (consistency priority) to AP (availability priority) +- Tiered Strategy: Set consistency priority levels for tables +- Degradation Mechanism: Prioritize service recovery for tables with lower consistency requirements, rather than waiting for complete data recovery +3.4 Version Upgrade Practice +3.x Version Implementation +- Validation Strategy: Build independent test clusters for functionality and stability verification +- Gradual Rollout: Progressively deploy community 3.x version in production environment +- Lessons Learned: Challenges and best practices during version upgrade +3.5 Data Structuring Transformation +HBase Table Schema Definition +- Pain Points: Address the issue that native HBase only has column family-level constraints, lacking column-level type definitions +- Solution Design: Define field types and field-to-column-family/column mapping relationships +- Benefits: + - Unified table structure management for business teams + - Automatic serialization/deserialization + - Foundation for SQL engine capabilities +3.6 SQL Engine Capabilities +SQL Semantics Implementation Based on Structuring +- Technical Foundation: Built upon data structuring achievements +- Core Capabilities: Implement standard SQL query semantics +- Usability Improvements: + - More convenient console queries + - Friendlier offline job data access + - Lower barrier to entry and improved readability +3.7 HBase Containerization +- Deployment Method: Deploy containerized Master and RegionServer processes based on Kubernetes +- Capability Enhancements: + - Improved elastic scaling + - Higher resource utilization + - Enhanced operation automation +3.8 HBase Storage-Compute Separation Architecture +- Technical Solution: Implement storage-compute separation based on JuiceFS + Object Storage +- Cost Reduction Impact: Significantly reduce storage costs +- Key Technologies: + - Smooth file system migration solution + - Interface adaptation and performance tuning + - Ensure seamless migration for business services + +--- +4. Summary +Through the above series of technical evolution, Xiaomi HBase system has achieved dual goals of performance improvement and cost optimization while ensuring business stability, providing practical reference for the evolution of large-scale distributed storage systems. + +### Speakers: + + +
+ +Yuting sun: Storage R&D Engineer + +Xiaomi software R&D engineer, mainly responsible for HBase development + + +
+ +Yiming Gong: Xiaomi, SDE + +HBase developer at xiaomi \ No newline at end of file diff --git a/content/sessions/datastorage-1211771.zh.md b/content/sessions/datastorage-1211771.zh.md new file mode 100644 index 000000000..2490daae1 --- /dev/null +++ b/content/sessions/datastorage-1211771.zh.md @@ -0,0 +1,81 @@ +--- +title: "小米 HBase 功能迭代与架构演进实践" +date: "2026-08-07T15:45:00" +track: "datastorage" +presenters: "Yuting sun, Yiming Gong" +stype: "中文演讲" +room: "万寿山会议室" +--- + +2. 描述： +本次演讲将介绍小米集团 HBase 系统的功能迭代与部署架构演进，以及在稳定性、性能和成本优化方面所做的工作，包括字典压缩、数据结构化、SQL 引擎能力、容器化以及存算分离。 +3. 演讲大纲 +3.1 数据压缩与存储优化 +字典压缩技术 +- 核心原理：基于高频字节序列构建辅助压缩结构 +- 技术特点：压缩与解压使用同一字典，尤其适用于具有重复模式的结构化数据 +- 实践成效：显著降低存储成本，提升 I/O 效率 +3.2 可观测性与治理能力 +表血缘（Lineage）特性 +- 功能：实现轻量级的表血缘采集能力 +- 采集信息：客户端版本、IP 地址、Kerberos 账号等关键元数据 +- 应用价值：支持一键血缘查询，用于数据治理与问题排查 +3.3 一致性与可用性的权衡 +HBase AP 实践探索 +- 定位转变：从 CP（一致性优先）向 AP（可用性优先）的探索 +- 分级策略：为表设置一致性优先级 +- 降级机制：优先恢复一致性要求较低的表的服务，而非等待数据完全恢复 +3.4 版本升级实践 +3.x 版本落地 +- 验证策略：搭建独立测试集群进行功能与稳定性验证 +- 逐步推广：在生产环境中渐进式部署社区 3.x 版本 +- 经验教训：版本升级过程中的挑战与最佳实践 +3.5 数据结构化改造 +HBase 表 Schema 定义 +- 痛点：解决原生 HBase 仅有列族级约束、缺少列级类型定义的问题 +- 方案设计：定义字段类型以及字段到列族/列的映射关系 +- 收益： + - 为业务团队提供统一的表结构管理 + - 自动序列化/反序列化 + - 为 SQL 引擎能力奠定基础 +3.6 SQL 引擎能力 +基于结构化的 SQL 语义实现 +- 技术基础：建立在数据结构化成果之上 +- 核心能力：实现标准 SQL 查询语义 +- 易用性提升： + - 更便捷的控制台查询 + - 更友好的离线作业数据访问 + - 更低的使用门槛与更好的可读性 +3.7 HBase 容器化 +- 部署方式：基于 Kubernetes 部署容器化的 Master 和 RegionServer 进程 +- 能力提升： + - 更好的弹性伸缩 + - 更高的资源利用率 + - 更强的运维自动化 +3.8 HBase 存算分离架构 +- 技术方案：基于 JuiceFS + 对象存储实现存算分离 +- 降本成效：显著降低存储成本 +- 关键技术： + - 平滑的文件系统迁移方案 + - 接口适配与性能调优 + - 确保业务服务无缝迁移 + +--- +4. 总结 +通过上述一系列技术演进，小米 HBase 系统在保障业务稳定的同时，达成了性能提升与成本优化的双重目标，为大规模分布式存储系统的演进提供了实践参考。 + +### 讲师: + + +
+ +Yuting sun：存储研发工程师 + +小米软件研发工程师，主要负责 HBase 开发。 + + +
+ +Yiming Gong：小米，SDE + +小米 HBase 开发者。 \ No newline at end of file diff --git a/content/sessions/datastorage-1212205.md b/content/sessions/datastorage-1212205.md new file mode 100644 index 000000000..ea4271d06 --- /dev/null +++ b/content/sessions/datastorage-1212205.md @@ -0,0 +1,35 @@ +--- +title: "Scaling Apache Ozone at Didi: Operating 100PB+ Storage and Billions of Files in Production" +date: "2026-08-07T17:15:00" +track: "datastorage" +presenters: "Shilun Fan, Ming Wei, Hongbing Wang" +stype: "Chinese Session" +room: "Mtn WanShou Hall" +--- + +As Didi’s unstructured data grew rapidly, HDFS metadata pressure became a major scalability challenge. Apache Ozone was introduced as the next-generation storage engine to address these limitations, and it has now been running in production at Didi for more than two years, supporting over 100 PB of data and tens of billions of files. + +This session gives an introductory overview of Didi’s Ozone journey, including why Ozone was chosen, how the platform evolved in production, and what engineering work was required to improve scalability, performance, and reliability. We will share practical experience in multi-cluster routing, follower-read optimization for S3 workloads, read performance tuning, erasure coding adoption, large-scale deletion and migration, observability, and availability improvements. The session will also summarize lessons learned from operating Ozone at scale and contributing improvements back to the Apache Ozone community. + +### Speakers: + + +
+ +Shilun Fan: Storage Engineer at DiDi, Apache Ozone Contributor + +Shilun Fan is an Apache Hadoop PMC Member, an Apache Auron Committer, and an Apache Ozone Contributor. He works on large-scale distributed storage systems and data infrastructure at DiDi. He has been deeply involved in the production deployment and optimization of Apache Ozone, with a focus on scalability, performance tuning, erasure coding, and reliability. His work focuses on building scalable, reliable, and cost-efficient storage platforms for massive data workloads. + + +
+ +Ming Wei: Storage Engineer at DiDi + +Apache Ozone Contributor + + +
+ +Hongbing Wang: Storage Engineer at DiDi + +Apache Ozone Committer \ No newline at end of file diff --git a/content/sessions/datastorage-1212205.zh.md b/content/sessions/datastorage-1212205.zh.md new file mode 100644 index 000000000..b0723125b --- /dev/null +++ b/content/sessions/datastorage-1212205.zh.md @@ -0,0 +1,35 @@ +--- +title: "滴滴 Apache Ozone 规模化实践：生产环境运营 100PB+ 存储与数十亿文件" +date: "2026-08-07T17:15:00" +track: "datastorage" +presenters: "Shilun Fan, Ming Wei, Hongbing Wang" +stype: "中文演讲" +room: "万寿山会议室" +--- + +随着滴滴非结构化数据快速增长，HDFS 元数据压力成为主要的可扩展性挑战。滴滴引入 Apache Ozone 作为下一代存储引擎来突破这些瓶颈，目前 Ozone 已在滴滴稳定运行两年多，支撑超过 100 PB 数据和数百亿文件。 + +本次演讲将对滴滴的 Ozone 历程做入门级概览，包括当初为何选择 Ozone、平台在生产环境中的演进，以及为提升可扩展性、性能和可靠性所做的工程工作。我们将分享在多集群路由、面向 S3 工作负载的 follower-read 优化、读性能调优、纠删码应用、大规模删除与迁移、可观测性以及可用性提升等方面的实践经验。演讲还会总结大规模运营 Ozone 的经验教训，以及把改进回馈给 Apache Ozone 社区的过程。 + +### 讲师: + + +
+ +Shilun Fan：滴滴存储工程师，Apache Ozone 贡献者 + +Shilun Fan 是 Apache Hadoop PMC 成员、Apache Auron Committer，以及 Apache Ozone 贡献者。他在滴滴从事大规模分布式存储系统与数据基础设施的工作。他深度参与了 Apache Ozone 的生产部署与优化，重点关注可扩展性、性能调优、纠删码与可靠性。他的工作重心是为海量数据工作负载构建可扩展、可靠且高性价比的存储平台。 + + +
+ +Ming Wei：滴滴存储工程师 + +Apache Ozone 贡献者 + + +
+ +Hongbing Wang：滴滴存储工程师 + +Apache Ozone Committer \ No newline at end of file diff --git a/content/sessions/datastorage-1212544.md b/content/sessions/datastorage-1212544.md new file mode 100644 index 000000000..045029348 --- /dev/null +++ b/content/sessions/datastorage-1212544.md @@ -0,0 +1,22 @@ +--- +title: "From Vision to Production: How Apache Fluss Defines the Storage Layer for Real-Time Data Lakehouses" +date: "2026-08-07T15:00:00" +track: "datastorage" +presenters: "Yang Wang" +stype: "Chinese Session" +room: "Mtn WanShou Hall" +--- + +Last year at Community Over Code Asia, we introduced Apache Fluss and its vision of a dedicated storage layer for real-time lakehouses. Several releases later, that vision has been validated in production at Alibaba's Double 11 scale. + +In this talk, we present what we have delivered across three themes. First, pushing compute into storage: the Aggregation Merge Engine pushes real-time aggregation — including RoaringBitmap UV deduplication — into the storage layer with built-in exactly-once semantics, and log filter pushdown moves predicate evaluation from compute to storage, enabling Flink to offload aggregation, join, and wide-table state entirely to storage. Second, ecosystem and data model evolution: secondary indexes for flexible query patterns, Change Data Feed, complex data types, zero-copy schema evolution, and multi-language clients (Rust, Python, C++). Third, enterprise-grade foundation: goal-driven rebalance, zero-downtime rolling upgrades, fast Coordinator recovery, SASL authentication with three-level ACL, and deep RocksDB diagnostics. + + +### Speakers: + + +
+ +Yang Wang: Staff Engineer at Alibaba Cloud , HangZhou, Apache Fluss Core Developer + +Yang Wang is a Staff Engineer at Alibaba Cloud and a core development team member of Apache Fluss, a streaming storage system for real-time analytics. He focuses on storage engine internals — including the KV store (RocksDB integration, performance tuning, monitoring, and memory management), aggregation merge engine, secondary index support, and Flink connector enhancements such as filter pushdown and lookup joins. He is an active contributor to the Apache Fluss open-source community. \ No newline at end of file diff --git a/content/sessions/datastorage-1212544.zh.md b/content/sessions/datastorage-1212544.zh.md new file mode 100644 index 000000000..720ff98f3 --- /dev/null +++ b/content/sessions/datastorage-1212544.zh.md @@ -0,0 +1,22 @@ +--- +title: "从愿景到生产：Apache Fluss 如何定义实时数据湖仓的存储层" +date: "2026-08-07T15:00:00" +track: "datastorage" +presenters: "Yang Wang" +stype: "中文演讲" +room: "万寿山会议室" +--- + +去年的 Community Over Code Asia 上，我们介绍了 Apache Fluss，以及它为实时湖仓打造专用存储层的愿景。经过几个版本的迭代，这一愿景已在阿里巴巴双十一规模的生产中得到验证。 + +在本次演讲中，我们将围绕三个主题介绍已交付的成果。第一，将计算下推到存储：聚合合并引擎（Aggregation Merge Engine）把实时聚合——包括基于 RoaringBitmap 的 UV 去重——下推到存储层，并内置 exactly-once 语义；日志过滤下推（log filter pushdown）则把谓词求值从计算侧移到存储侧，使 Flink 能够把聚合、join 和宽表状态完全卸载到存储。第二，生态与数据模型演进：用于灵活查询模式的二级索引、Change Data Feed、复杂数据类型、零拷贝模式演进，以及多语言客户端（Rust、Python、C++）。第三，企业级基础设施：目标驱动的 rebalance、零停机滚动升级、快速的 Coordinator 恢复、带三级 ACL 的 SASL 认证，以及深度的 RocksDB 诊断。 + + +### 讲师: + + +
+ +Yang Wang：阿里云高级主任工程师，杭州，Apache Fluss 核心开发者 + +Yang Wang 是阿里云高级主任工程师，也是面向实时分析的流式存储系统 Apache Fluss 的核心开发团队成员。他专注于存储引擎内部实现——包括 KV 存储（RocksDB 集成、性能调优、监控与内存管理）、聚合合并引擎、二级索引支持，以及 Flink connector 的增强（如过滤下推、lookup join）。他是 Apache Fluss 开源社区的活跃贡献者。 \ No newline at end of file diff --git a/content/sessions/datastorage-1213149.md b/content/sessions/datastorage-1213149.md new file mode 100644 index 000000000..c0a022121 --- /dev/null +++ b/content/sessions/datastorage-1213149.md @@ -0,0 +1,33 @@ +--- +title: "When the Optimizer Lies: Debugging Cross-Slice Execution in Apache Cloudberry" +date: "2026-08-08T16:45:00" +track: "datastorage" +presenters: "Alena Rybakina" +stype: "English Session" +room: "Mtn WanShou Hall" +--- + +In Apache Cloudberry, the optimizer, planner, and executor operate with different abstractions of a query. When these abstractions diverge, even simple queries can trigger severe failures. +This talk presents a real bug involving Shared Scans over Common Table Expressions (CTEs) on replicated tables. When such CTEs were referenced multiple times as scalar subqueries, queries could hang indefinitely or fail with errors like "could not open existing temporary file". +The root cause was a mismatch between logical and physical execution: scalar SubPlans hid dependencies from the optimizer, leading to incorrect slice assignment and broken producer–consumer locality. +In this talk, I will walk through several alternative fixes and explain how each of them attempts to address the problem. We will dive into the optimizer internals to understand why the bug happens in the first place, how different fixes change the execution topology, and why some of them - despite working - cannot be accepted as a final solution. +Some fixes introduce subtle regressions, change plan shapes, or force fallback to the Postgres optimizer, highlighting the difficulty of making local improvements without breaking global behavior. +This talk shows how a seemingly simple query pattern can expose deep inconsistencies between optimizer and execution - and why fixing such issues often requires navigating trade-offs rather than finding a single “correct” solution. + +### Speakers: + + +
+ +Alena Rybakina: Software developer, Yandex Cloud + +Work Experience +Over 5 years of professional experience in database systems, including 5 years at Postgres Professional. +Main Interests +Query optimization, MVCC internals, and database statistics. +PostgreSQL Experience +More than 4 years of hands-on experience working on PostgreSQL internals, performance improvements, and optimizer-related features. +Main Projects and Contributions +Development of optimization-related extensions and features, including AQO (Adaptive Query Optimization) and Self-Join Elimination +Experimental and production-oriented improvements to the PostgreSQL query planner +Contributions to PostgreSQL core, including the OR-to-ANY and VALUES-to-ANY transformations in the optimizer \ No newline at end of file diff --git a/content/sessions/datastorage-1213149.zh.md b/content/sessions/datastorage-1213149.zh.md new file mode 100644 index 000000000..cdbef4fc8 --- /dev/null +++ b/content/sessions/datastorage-1213149.zh.md @@ -0,0 +1,33 @@ +--- +title: "当优化器说谎时：调试 Apache Cloudberry 中的跨 Slice 执行" +date: "2026-08-08T16:45:00" +track: "datastorage" +presenters: "Alena Rybakina" +stype: "英文演讲" +room: "万寿山会议室" +--- + +在 Apache Cloudberry 中，优化器、规划器（planner）和执行器对同一条查询持有不同的抽象。当这些抽象之间出现分歧时，即使是简单的查询也可能引发严重的故障。 +本次演讲分享一个涉及复制表（replicated table）上公共表表达式（CTE）共享扫描（Shared Scan）的真实 bug。当此类 CTE 被作为标量子查询多次引用时，查询可能无限期挂起，或报出"could not open existing temporary file"之类的错误。 +根本原因在于逻辑执行与物理执行之间的错配：标量 SubPlan 向优化器隐藏了依赖关系，导致 slice 分配不正确，破坏了生产者—消费者的局部性。 +在本次演讲中，我将梳理几种可选的修复方案，并解释每种方案是如何尝试解决这一问题的。我们会深入优化器内部，理解这个 bug 为何会产生、不同修复方案如何改变执行拓扑，以及为什么其中一些方案——尽管可行——却无法作为最终方案被接受。 +有些修复会引入细微的回归、改变计划形态，或被迫回退到 Postgres 优化器，这凸显了在不破坏全局行为的前提下做局部改进的困难。 +本次演讲展示了看似简单的查询模式如何暴露优化器与执行器之间深层的不一致——以及为什么修复这类问题往往需要在权衡中抉择，而非寻找一个唯一"正确"的答案。 + +### 讲师: + + +
+ +Alena Rybakina：软件开发工程师，Yandex Cloud + +工作经历 +拥有 5 年以上数据库系统专业经验，其中包括在 Postgres Professional 工作的 5 年。 +主要兴趣 +查询优化、MVCC 内部机制与数据库统计信息。 +PostgreSQL 经验 +拥有 4 年以上从事 PostgreSQL 内部实现、性能改进及优化器相关功能的一手经验。 +主要项目与贡献 +开发与优化相关的扩展和功能，包括 AQO（自适应查询优化，Adaptive Query Optimization）和自连接消除（Self-Join Elimination）。 +对 PostgreSQL 查询规划器的实验性及面向生产的改进。 +对 PostgreSQL 核心的贡献，包括优化器中的 OR-to-ANY 和 VALUES-to-ANY 转换。 \ No newline at end of file diff --git a/content/sessions/datastorage-1213232.md b/content/sessions/datastorage-1213232.md new file mode 100644 index 000000000..2c713fbe6 --- /dev/null +++ b/content/sessions/datastorage-1213232.md @@ -0,0 +1,21 @@ +--- +title: "Everything boil down to SQL: Shift DataFusion Optimizations to Another Non-Generic Computing System" +date: "2026-08-07T14:30:00" +track: "datastorage" +presenters: "Ruihang Xia" +stype: "Chinese Session" +room: "Mtn WanShou Hall" +--- + +On the way to re-implement PromQL (the query language behind Prometheus) on top of Apache DataFusion, we found many similar optimization patterns between these two totally different languages. Some share a similar idea, some can be mixed together, and some are fed back to DataFusion. We further confirmed that almost all different-looking computing systems, no matter what their user interface looks like, are composed of several generic primitives. In this session, we'll first discuss those primitives and their optimizations. Then talk about high-level philosophies of how to keep things simple and unified while making optimizations for different systems and purposes, with lots of famous systems as examples. + +### Speakers: + + +
+ +Ruihang Xia: Software Architect at Greptime Inc. + +Software Architect at Greptime +Apache DataFusion PMC, Apache Arrow Committer, Apache HoraeDB PPMC +https://github.com/waynexia \ No newline at end of file diff --git a/content/sessions/datastorage-1213232.zh.md b/content/sessions/datastorage-1213232.zh.md new file mode 100644 index 000000000..4123e8ff7 --- /dev/null +++ b/content/sessions/datastorage-1213232.zh.md @@ -0,0 +1,21 @@ +--- +title: "万法归 SQL：把 DataFusion 的优化迁移到另一个非通用计算系统" +date: "2026-08-07T14:30:00" +track: "datastorage" +presenters: "Ruihang Xia" +stype: "中文演讲" +room: "万寿山会议室" +--- + +在基于 Apache DataFusion 重新实现 PromQL（Prometheus 背后的查询语言）的过程中，我们在这两种截然不同的语言之间发现了许多相似的优化模式。有些思路相近，有些可以融合在一起，还有一些被反哺回了 DataFusion。我们进一步确信：几乎所有看起来各不相同的计算系统，无论其用户界面长什么样，本质上都由若干通用的原语（primitive）组合而成。在本次演讲中，我们会先讨论这些原语及其优化，随后谈谈如何在为不同系统和不同目的做优化时，保持简洁与统一的高层哲学，并以大量知名系统作为案例。 + +### 讲师: + + +
+ +Ruihang Xia：Greptime 软件架构师 + +Greptime 软件架构师 +Apache DataFusion PMC、Apache Arrow Committer、Apache HoraeDB PPMC +https://github.com/waynexia \ No newline at end of file diff --git a/content/sessions/datastorage-1213677.md b/content/sessions/datastorage-1213677.md new file mode 100644 index 000000000..850b3838f --- /dev/null +++ b/content/sessions/datastorage-1213677.md @@ -0,0 +1,28 @@ +--- +title: "Scaling Beyond the Exabyte: Lessons from Hadoop Integration and Operations at LY Corporation" +date: "2026-08-08T14:30:00" +track: "datastorage" +presenters: "Akira Ajisaka, Kai Hirayama" +stype: "English Session" +room: "Mtn WanShou Hall" +--- + +Following the merger of LINE and Yahoo! JAPAN, we faced the monumental task of enabling seamless data access between two massive, completely independent Hadoop ecosystems—each with its own isolated networks and disparate account systems. This session provides a deep dive into the technical strategies we used to overcome these architectural barriers and achieve cross-cluster interoperability. + +Furthermore, we will explore our journey of scaling and performance tuning using NameNode Federation (including Router-based Federation) and Observer NameNodes to push past the inherent limits of a single NameNode. We will share the real-world operational hurdles we encountered and the specific tuning required to maintain stability in one of the world's largest Hadoop deployments. Finally, we will discuss our experience contributing patches back to the OSS community to ensure these features are battle-tested for enterprise-scale requirements. + +### Speakers: + + +
+ +Akira Ajisaka: Principal Engineer, LY Corporation + +Akira Ajisaka is a Principal Engineer at LY Corporation, with over 10 years of expertise in big data technologies across both on-premises and public cloud environments. Apache Member, Apache Hadoop PMC member, and Apache Kyuubi PMC member. + + +
+ +Kai Hirayama: Data Engineer, LY Corporation + +Data Platform Engineer working on Hadoop ecosystem, including YARN and HDFS. Experienced in operating and troubleshooting large-scale distributed systems. \ No newline at end of file diff --git a/content/sessions/datastorage-1213677.zh.md b/content/sessions/datastorage-1213677.zh.md new file mode 100644 index 000000000..11b809cd0 --- /dev/null +++ b/content/sessions/datastorage-1213677.zh.md @@ -0,0 +1,28 @@ +--- +title: "跨越 EB 级规模：LY Corporation 的 Hadoop 集成与运营经验" +date: "2026-08-08T14:30:00" +track: "datastorage" +presenters: "Akira Ajisaka, Kai Hirayama" +stype: "英文演讲" +room: "万寿山会议室" +--- + +在 LINE 与 Yahoo! JAPAN 合并之后，我们面临一项艰巨的任务：在两个庞大且完全独立、各自拥有孤立网络和不同账户体系的 Hadoop 生态之间，实现无缝的数据访问。本次演讲将深入剖析我们用来突破这些架构壁垒、实现跨集群互操作的技术策略。 + +此外，我们将分享利用 NameNode Federation（包括基于 Router 的 Federation）和 Observer NameNode 进行扩展与性能调优的历程，以突破单个 NameNode 的固有上限。我们会分享所遇到的真实运维难题，以及为在这个全球最大规模的 Hadoop 部署之一中维持稳定性所需的针对性调优。最后，我们还会讨论把补丁回馈给开源社区的经验，以确保这些特性经过企业级要求的实战检验。 + +### 讲师: + + +
+ +Akira Ajisaka：LY Corporation 首席工程师 + +Akira Ajisaka 是 LY Corporation 的首席工程师，在本地和公有云环境下的大数据技术领域拥有超过 10 年的专业经验。他是 Apache 成员、Apache Hadoop PMC 成员，以及 Apache Kyuubi PMC 成员。 + + +
+ +Kai Hirayama：LY Corporation 数据工程师 + +数据平台工程师，从事包括 YARN 和 HDFS 在内的 Hadoop 生态工作，在大规模分布式系统的运维与排障方面经验丰富。 \ No newline at end of file diff --git a/content/sessions/datastorage-1214223.md b/content/sessions/datastorage-1214223.md new file mode 100644 index 000000000..0ec00b679 --- /dev/null +++ b/content/sessions/datastorage-1214223.md @@ -0,0 +1,23 @@ +--- +title: "Unifying Lance Metadata Management with Gravitino" +date: "2026-08-08T15:45:00" +track: "datastorage" +presenters: "Qi Yu" +stype: "Chinese Session" +room: "Mtn WanShou Hall" +--- + +With the rise of AI and vector-based workloads, Lance is gaining traction as a high-performance data format for analytical and ML use cases. However, managing Lance datasets across systems remains challenging due to the lack of a unified metadata layer. + +In this talk, we introduce how Apache Gravitino enables centralized metadata management for Lance through its Lance REST service integration. We will walk through how Lance datasets can be modeled using a unified catalog abstraction, making them easier to manage, discover, and access across different engines. + +We will also share the design considerations behind this integration, demonstrate how to create and access Lance datasets via Gravitino, and discuss how this approach helps build a more consistent and extensible data platform for both traditional analytics and emerging AI workload. + +### Speakers: + + +
+ +Qi Yu: Data infrastructure engineer, Apache Gravitino PMC + +Engineer at Datastato and PMC member of Apache Gravitino, working on big data infrastructure, metadata management, and database systems, with a focus on scalable data platform development. \ No newline at end of file diff --git a/content/sessions/datastorage-1214223.zh.md b/content/sessions/datastorage-1214223.zh.md new file mode 100644 index 000000000..e06594073 --- /dev/null +++ b/content/sessions/datastorage-1214223.zh.md @@ -0,0 +1,23 @@ +--- +title: "用 Gravitino 统一 Lance 的元数据管理" +date: "2026-08-08T15:45:00" +track: "datastorage" +presenters: "Qi Yu" +stype: "中文演讲" +room: "万寿山会议室" +--- + +随着 AI 和基于向量的工作负载兴起，Lance 作为面向分析与机器学习场景的高性能数据格式，正受到越来越多的关注。然而，由于缺少统一的元数据层，跨系统管理 Lance 数据集仍然充满挑战。 + +在本次演讲中，我们将介绍 Apache Gravitino 如何通过其 Lance REST 服务集成，实现对 Lance 的集中化元数据管理。我们将演示如何用统一的目录抽象来建模 Lance 数据集，使其更易于在不同引擎间被管理、发现和访问。 + +我们还会分享这次集成背后的设计考量，演示如何通过 Gravitino 创建并访问 Lance 数据集，并讨论这种做法如何帮助构建一个更一致、更具扩展性的数据平台，以同时服务传统分析和新涌现的 AI 工作负载。 + +### 讲师: + + +
+ +Qi Yu：数据基础设施工程师，Apache Gravitino PMC + +Datastrato 工程师、Apache Gravitino PMC 成员，从事大数据基础设施、元数据管理与数据库系统的工作，专注于可扩展数据平台的研发。 \ No newline at end of file diff --git a/content/sessions/datastorage-1214278.md b/content/sessions/datastorage-1214278.md new file mode 100644 index 000000000..3cda1ccec --- /dev/null +++ b/content/sessions/datastorage-1214278.md @@ -0,0 +1,33 @@ +--- +title: "Paimon-cpp: Bringing Native High-Performance Data Lakehouse Access to the Apache Ecosystem" +date: "2026-08-08T15:00:00" +track: "datastorage" +presenters: "Xinyu Liu" +stype: "Chinese Session" +room: "Mtn WanShou Hall" +--- + +Apache Paimon has become one of the most actively developed data lakehouse formats in the Apache ecosystem. As native query engines increasingly demand direct access to lakehouse data without JVM overhead, we are excited to introduce Paimon-cpp — a high-performance, ground-up C++ implementation of the Apache Paimon format, in the process of being contributed to the Apache Paimon community. + +In this session, we will walk through: + +- **Why Paimon-cpp**: The motivation behind building a native C++ implementation — bridging the gap between JVM-based data lake ecosystems and the growing world of native query engines, while maintaining full wire-compatibility with Java Paimon. + +- **Format & Features**: Paimon-cpp provides full read/write/compaction support for both append tables and primary key tables, including Merge-On-Read (MOR) and Deletion Vector (DV) scenarios. It also supports schema evolution, predicate push-down, AI-oriented data evolution mode, and various index types (e.g., vector search, full text search). + +- **Performance Optimizations**: Shallow data copy data exchange, file prefetching for read-ahead, multi-threaded & asynchronous row-columnar conversion for primary key tables, and Blob I/O optimizations. + +- **Architecture & Extensibility**: Built on the Apache Arrow Columnar In-Memory Format, with a modular, plugin-oriented architecture providing well-defined interfaces for file systems, file formats, memory management, executors, and observability — designed to be easily embedded into any native engine. + +- **Best Practices & Getting Started**: Practical guidance on configuring write/read/compaction pipelines, tuning for production workloads, and integrating Paimon-cpp into your own engine. + +Paimon-cpp is being actively contributed to the Apache Paimon project and is expected to be part of the official Apache repository by the time of this talk. We warmly welcome developers, engine builders, and data enthusiasts to try it out, report issues, and contribute — whether it's new features, bug fixes, documentation, or integration with your favorite native engine. Let's build the native data lakehouse ecosystem together! + +### Speakers: + + +
+ +Xinyu Liu: Senior Software Engineer at Alibaba | Maintainer of Paimon-cpp | 3 Years with Paimon + +I am a Senior Software Engineer on the Storage Service team at Alibaba. I have been working on Paimon-cpp for 3 years, currently serving as the maintainer of the Paimon-cpp open-source library and leading the adoption of Paimon-cpp across Alibaba's internal data infrastructure. Prior to Paimon, I was responsible for Khronos, a time-series database that powers the monitoring system storage and retrieval for Alibaba Group. My experience spans storage engine design, high-performance data formats, and systems-level optimization. I am passionate about open-source collaboration and look forward to building the native data lakehouse ecosystem together with the Apache community. \ No newline at end of file diff --git a/content/sessions/datastorage-1214278.zh.md b/content/sessions/datastorage-1214278.zh.md new file mode 100644 index 000000000..9d1547ff1 --- /dev/null +++ b/content/sessions/datastorage-1214278.zh.md @@ -0,0 +1,33 @@ +--- +title: "Paimon-cpp：为 Apache 生态带来原生高性能的湖仓访问" +date: "2026-08-08T15:00:00" +track: "datastorage" +presenters: "Xinyu Liu" +stype: "中文演讲" +room: "万寿山会议室" +--- + +Apache Paimon 已成为 Apache 生态中开发最活跃的湖仓格式之一。随着原生查询引擎越来越希望在不引入 JVM 开销的情况下直接访问湖仓数据，我们很高兴地介绍 Paimon-cpp——一套从头构建、高性能的 Apache Paimon 格式 C++ 实现，目前正贡献给 Apache Paimon 社区。 + +在本次演讲中，我们将涵盖： + +- **为什么是 Paimon-cpp**：构建原生 C++ 实现的动机——在基于 JVM 的数据湖生态与日益壮大的原生查询引擎世界之间架起桥梁，同时与 Java 版 Paimon 保持完全的线协议兼容（wire-compatibility）。 + +- **格式与特性**：Paimon-cpp 为 append 表和主键表都提供完整的读/写/compaction 支持，包括 Merge-On-Read（MOR）和 Deletion Vector（DV）场景。它还支持模式演进、谓词下推、面向 AI 的数据演进模式，以及多种索引类型（如向量检索、全文检索）。 + +- **性能优化**：浅拷贝数据交换、用于预读的文件预取（prefetch）、面向主键表的多线程异步行列转换，以及 Blob I/O 优化。 + +- **架构与可扩展性**：构建在 Apache Arrow 列式内存格式之上，采用模块化、插件化的架构，为文件系统、文件格式、内存管理、执行器和可观测性提供了定义良好的接口——设计上便于嵌入任意原生引擎。 + +- **最佳实践与上手指南**：关于配置读/写/compaction 流水线、面向生产负载的调优，以及把 Paimon-cpp 集成到你自有引擎的实操指南。 + +Paimon-cpp 正积极贡献给 Apache Paimon 项目，预计在本次演讲之时已成为 Apache 官方仓库的一部分。我们热忱欢迎开发者、引擎构建者和数据爱好者试用、反馈问题并贡献代码——无论是新功能、bug 修复、文档，还是与你钟爱的原生引擎的集成。让我们共同打造原生的湖仓生态！ + +### 讲师: + + +
+ +Xinyu Liu：阿里巴巴高级软件工程师 | Paimon-cpp Maintainer | 3 年 Paimon 经验 + +我是阿里巴巴存储服务团队的高级软件工程师。我从事 Paimon-cpp 工作已有 3 年，目前担任 Paimon-cpp 开源库的 maintainer，并负责推动 Paimon-cpp 在阿里巴巴内部数据基础设施中的落地。在 Paimon 之前，我负责 Khronos——一个支撑阿里巴巴集团监控系统存储与检索的时序数据库。我的经验涵盖存储引擎设计、高性能数据格式以及系统级优化。我对开源协作充满热情，期待与 Apache 社区共同建设原生湖仓生态。 \ No newline at end of file diff --git a/content/sessions/general-1194670.md b/content/sessions/general-1194670.md new file mode 100644 index 000000000..c4877a68e --- /dev/null +++ b/content/sessions/general-1194670.md @@ -0,0 +1,32 @@ +--- +title: "Accelerating Quantum Machine Learning: Building a GPU-Accelerated Data Plane in Apache Mahout" +date: "2026-08-08T16:15:00" +track: "general" +presenters: "Jie-Kai Chang, Guan-Hua Wen" +stype: "Chinese Session" +room: "Mtn BaiWang Hall" +--- + +Quantum machine learning (QML) holds great promise, but practical implementations often encounter a major bottleneck before quantum computation even begins. Classical feature vectors must first be prepared and encoded into quantum states. As data dimensionality and the number of qubits increase, this step can dominate end-to-end execution time and limit the scalability of QML workflows. + +In this session, we will introduce Quantum Data Plane (QDP), a GPU-accelerated data plane being developed in Apache Mahout to make quantum data encoding more efficient and scalable. QDP treats data encoding as a dedicated execution layer rather than a one-off preprocessing task. We will explain its motivation, overall architecture, integration with Apache Mahout, and how GPU parallelism can help address the encoding bottleneck. + +Drawing on our implementation experience, we will share key design choices and lessons learned from building QDP. We will discuss where GPU acceleration provides the greatest value, the limitations and trade-offs of this approach, and how the project may evolve through open-source collaboration. + +Attendees will leave with a practical understanding of data preparation challenges in QML and a broader perspective on how Apache Mahout can connect classical machine learning infrastructure with emerging quantum computing workflows. + +### Speakers: + + +
+ +Jie-Kai Chang: Apache Mahout PMC Member + +Jie-Kai Chang is actively involved in open-source communities and project governance, with a focus on high-performance computing, GPU acceleration, and machine learning infrastructure. He also participates in the Ray and KubeRay communities, exploring distributed computing, cloud-native AI, and AI workload management on Kubernetes. His current work addresses data-encoding bottlenecks in quantum machine learning and advances the design and implementation of Apache Mahout's Quantum Data Plane (QDP). + + +
+ +Guan-Hua Wen: Apache Mahout Committer + +Research and Development Intern at Microsoft \ No newline at end of file diff --git a/content/sessions/general-1194670.zh.md b/content/sessions/general-1194670.zh.md new file mode 100644 index 000000000..e3cc517a0 --- /dev/null +++ b/content/sessions/general-1194670.zh.md @@ -0,0 +1,32 @@ +--- +title: "加速量子机器学习：在 Apache Mahout 中构建 GPU 加速数据平面" +date: "2026-08-08T16:15:00" +track: "general" +presenters: "张杰凯, 温冠华" +stype: "中文演讲" +room: "百望山会议室" +--- + +量子机器学习（QML）虽然前景广阔，但在实际应用中，往往会在量子计算开始之前遇到一个主要瓶颈。经典特征向量必须先经过处理并编码成量子态。随着数据维度和量子比特数量的增加，这一步骤可能占据端到端执行时间的大部分，并限制 QML 工作流的可扩展性。 + +本次演讲将介绍 Quantum Data Plane（QDP，量子数据平面）。这是 Apache Mahout 中正在开发的 GPU 加速数据平面，旨在让量子数据编码更加高效且更具可扩展性。QDP 将数据编码视为一个专门的执行层，而不是一次性的预处理任务。我们将介绍其设计动机、整体架构、与 Apache Mahout 的集成方式，以及 GPU 并行计算如何帮助解决数据编码瓶颈。 + +结合实际开发经验，我们将分享构建 QDP 过程中的关键设计选择与经验总结，并讨论 GPU 加速最能发挥价值的场景、这种方法的局限与取舍，以及项目未来如何通过开源协作持续发展。 + +听众将了解 QML 中数据准备所面临的实际挑战，并从更广泛的视角认识 Apache Mahout 如何连接经典机器学习基础设施与新兴的量子计算工作流。 + +### 讲师: + + +
+ +张杰凯：Apache Mahout 项目管理委员会成员 + +张杰凯长期参与开源社区与项目治理，关注高性能计算、GPU 加速及机器学习基础设施。他也参与 Ray 与 KubeRay 社区，探索分布式计算、云原生 AI，以及 Kubernetes 上的 AI 工作负载管理。目前致力于解决量子机器学习中的数据编码瓶颈，并推动 Apache Mahout Quantum Data Plane（QDP）的设计与实现。 + + +
+ +温冠华：Apache Mahout Committer + +微软研发实习生 \ No newline at end of file diff --git a/content/sessions/general-1205364.md b/content/sessions/general-1205364.md new file mode 100644 index 000000000..d175f2e0c --- /dev/null +++ b/content/sessions/general-1205364.md @@ -0,0 +1,23 @@ +--- +title: "Apache CloudStack: The Open Source Cloud Platform Powering Modern Infrastructure" +date: "2026-08-08T14:30:00" +track: "general" +presenters: "Abhishek Kumar" +stype: "English Session" +room: "Mtn BaiWang Hall" +--- + +Apache CloudStack is a mature open-source cloud management platform used by service providers and enterprises to build scalable Infrastructure-as-a-Service (IaaS) clouds. + +In this session, we will introduce the architecture and core concepts behind CloudStack and explore how it simplifies the orchestration of compute, storage, and networking across diverse infrastructure environments. We will also discuss why organizations choose CloudStack as a powerful alternative to other cloud management platforms. + +The talk will highlight some of the latest innovations and key capabilities in the CloudStack ecosystem, demonstrating how the platform continues to evolve to meet modern infrastructure needs. + +### Speakers: + + +
+ +Abhishek Kumar: Apache CloudStack PMC member + +Abhishek Kumar is a Software Engineer at ShapeBlue, where he works on Apache CloudStack and cloud infrastructure solutions. He is an active contributor to the CloudStack project, focusing on platform extensibility, integrations, and improvements to cloud orchestration. His interests include virtualization, large-scale infrastructure automation, and advancing open-source cloud technologies. \ No newline at end of file diff --git a/content/sessions/general-1205364.zh.md b/content/sessions/general-1205364.zh.md new file mode 100644 index 000000000..734218d31 --- /dev/null +++ b/content/sessions/general-1205364.zh.md @@ -0,0 +1,23 @@ +--- +title: "Apache CloudStack：驱动现代基础设施的开源云平台" +date: "2026-08-08T14:30:00" +track: "general" +presenters: "Abhishek Kumar" +stype: "英文演讲" +room: "百望山会议室" +--- + +Apache CloudStack 是一个成熟的开源云管理平台，被众多服务提供商和企业用于构建可扩展的基础设施即服务（IaaS）云。 + +在本次演讲中，我们将介绍 CloudStack 背后的架构与核心概念，探讨它如何简化跨多种基础设施环境下的计算、存储和网络编排。我们还会讨论为什么众多组织会选择 CloudStack，将其作为其他云管理平台之外的有力替代方案。 + +演讲将重点介绍 CloudStack 生态中的一些最新创新与关键能力，展示该平台如何持续演进以满足现代基础设施的需求。 + +### 讲师: + + +
+ +Abhishek Kumar：Apache CloudStack PMC 成员 + +Abhishek Kumar 是 ShapeBlue 的软件工程师，从事 Apache CloudStack 及云基础设施解决方案的工作。他是 CloudStack 项目的活跃贡献者，专注于平台可扩展性、各类集成以及对云编排的改进。他的兴趣涵盖虚拟化、大规模基础设施自动化，以及推动开源云技术的发展。 \ No newline at end of file diff --git a/content/sessions/general-1213363.md b/content/sessions/general-1213363.md new file mode 100644 index 000000000..2902921ea --- /dev/null +++ b/content/sessions/general-1213363.md @@ -0,0 +1,42 @@ +--- +title: "Apache OpenMeetings: Enterprise Video Conferencing Meets AI Agents" +date: "2026-08-08T15:00:00" +track: "general" +presenters: "Wentao Liu, Zeguo Mao, Xianxiong Luo" +stype: "Chinese Session" +room: "Mtn BaiWang Hall" +--- + +Apache OpenMeetings is a mature, open-source video conferencing and collaboration platform trusted by enterprises and educational institutions worldwide. This session will provide a comprehensive overview of OpenMeetings' core capabilities including OAuth2 authentication, LDAP integration, virtual backgrounds, mobile applications, clustering for high availability, and WebService APIs for enterprise integration. This talk is designed for developers, system administrators, and decision-makers exploring self-hosted video conferencing solutions with modern AI enhancements. Key Takeaways of the talk are: Understand the enterprise‑grade capabilities of Apache OpenMeetings, Learn how to integrate OpenMeetings with existing identity providers and directory services, See how virtual backgrounds and mobile apps enhance the user experience, Discover clustering options for scaling to thousands of concurrent users, Extend OpenMeetings with custom WebService integrations, Witness a live demonstration of AI‑agent integration using OpenClaw.By the end of this session, you will understand OpenMeetings' architecture, deployment options, and how to extend it with AI agents to build next-generation collaboration platforms. + +### Speakers: + + +
+ +Wentao Liu: Manager of omfoss.com + +Wentao Liu has been a passionate contributor to the Apache Software Foundation for many years. With over two decades of experience in the Linux and open‑source ecosystem, he has played a pivotal role in advancing the OpenMeetings project—from core development and community building to real‑world enterprise adoption. + +His contributions include: +Delivering talks and workshops about Apache OpenMeetings at major open‑source conferences such as ApacheCon, FOSSASIA, FOSDEM, and CommunityOverCode, helping to evangelize the platform’s capabilities to a global audience. +Founding a company that builds commercial video‑conference appliances based on Apache OpenMeetings, demonstrating how the project can be tailored for secure, scalable, on‑premise deployments. +Driving the integration of modern authentication (OAuth2, LDAP), virtual backgrounds, mobile applications, and clustering features that make OpenMeetings ready for enterprise use. + + +
+ +Zeguo Mao: sophomore + +Zeguo Mao is a sophomore at Central South University of Forestry and Technology, majoring in computer science and Technology. With a strong foundation in C, C++, Java, and web technologies including SpringBoot, HTML/CSS, and JavaScript, Mao has cultivated a practical, project-driven approach to building interactive systems. His technical curiosity extends deeply into the realm of Artificial Intelligence, where he possesses a solid understanding of machine learning and deep learning algorithms. + +Mao's hands-on experience with intelligent devices, particularly the development of voice-interaction applications for Rokid smart glasses, gives him a unique perspective on human-computer collaboration. He led the "Universal Manual" project, creating a fully voice-controlled object recognition and intelligent explanation system that demonstrates the seamless integration of AR hardware with AI-driven context awareness. Additionally, as the principal investigator for the "Betel Nut Automatic Picking Robot" project, Mao was responsible for market research, feasibility analysis, and the comprehensive technical design of an autonomous system. + + +
+ +Xianxiong Luo: Organizer of GitCoffee Open Source Community + +Xianxiong is a seasoned technology leader with 17+ years of hands-on experience in software development, enterprise architecture, and technical team management across E-commerce, Ride-hailing, and Open Source ecosystems. Former Product Manager for the Membership Business Line at Better Life Group and former Architect & Technical Director at Guangcai Shenzhou Group. Currently serving as the Head of Yifei Open Source and Lead Maintainer of GitCoffee Open Source Community. + +Recognized as a 2024 & 2025 Outstanding Partner of the OpenTiny Open Source Community (under Huawei Cloud). Deeply focused on bridging open-source community building with enterprise-grade AI applications. Core technical expertise spans Java Backend Architecture, Full-Stack Development, and AI Agent Implementation. \ No newline at end of file diff --git a/content/sessions/general-1213363.zh.md b/content/sessions/general-1213363.zh.md new file mode 100644 index 000000000..f8302a6d2 --- /dev/null +++ b/content/sessions/general-1213363.zh.md @@ -0,0 +1,42 @@ +--- +title: "Apache OpenMeetings：企业级视频会议遇上 AI Agent" +date: "2026-08-08T15:00:00" +track: "general" +presenters: "Wentao Liu, Zeguo Mao, Xianxiong Luo" +stype: "中文演讲" +room: "百望山会议室" +--- + +Apache OpenMeetings 是一个成熟的开源视频会议与协作平台，受到全球企业与教育机构的信赖。本次演讲将全面介绍 OpenMeetings 的核心能力，包括 OAuth2 认证、LDAP 集成、虚拟背景、移动应用、面向高可用的集群，以及用于企业集成的 WebService API。本次演讲面向正在探索具备现代 AI 增强能力的自建视频会议方案的开发者、系统管理员和决策者。演讲的核心要点包括：了解 Apache OpenMeetings 的企业级能力；学习如何把 OpenMeetings 与现有的身份提供方和目录服务集成；了解虚拟背景和移动应用如何提升用户体验；了解用于扩展到数千并发用户的集群方案；通过自定义 WebService 集成扩展 OpenMeetings；以及亲眼见证使用 OpenClaw 进行 AI Agent 集成的现场演示。在本次演讲结束时，你将理解 OpenMeetings 的架构、部署选项，以及如何用 AI Agent 对其进行扩展，从而构建下一代协作平台。 + +### 讲师: + + +
+ +Wentao Liu：omfoss.com 负责人 + +Wentao Liu 多年来一直是 Apache 软件基金会充满热情的贡献者。他在 Linux 和开源生态拥有逾 20 年经验，在推动 OpenMeetings 项目发展方面发挥了关键作用——从核心开发、社区建设到真实的企业级落地。 + +他的贡献包括： +在 ApacheCon、FOSSASIA、FOSDEM、CommunityOverCode 等各大开源会议上发表关于 Apache OpenMeetings 的演讲和工作坊，帮助向全球受众推广该平台的能力。 +创办了一家基于 Apache OpenMeetings 构建商用视频会议硬件（appliance）的公司，展示了该项目如何为安全、可扩展的本地化部署进行定制。 +推动了现代认证（OAuth2、LDAP）、虚拟背景、移动应用和集群等特性的集成，让 OpenMeetings 为企业级使用做好了准备。 + + +
+ +Zeguo Mao：大二学生 + +Zeguo Mao 是中南林业科技大学计算机科学与技术专业的大二学生。他在 C、C++、Java 以及包括 SpringBoot、HTML/CSS、JavaScript 在内的 Web 技术方面拥有扎实基础，在构建交互式系统上形成了注重实践、以项目驱动的方法。他的技术好奇心深入人工智能领域，对机器学习和深度学习算法有扎实的理解。 + +Mao 在智能设备方面的一手经验——尤其是为 Rokid 智能眼镜开发语音交互应用——赋予了他关于人机协作的独特视角。他牵头了"通用手册"项目，打造了一套全语音控制的物体识别与智能讲解系统，展示了 AR 硬件与 AI 驱动的上下文感知的无缝集成。此外，作为"槟榔自动采摘机器人"项目的主要负责人，Mao 承担了市场调研、可行性分析，以及一套自主系统的全面技术设计工作。 + + +
+ +Xianxiong Luo：GitCoffee 开源社区组织者 + +Xianxiong 是一位资深技术负责人，在软件开发、企业架构和技术团队管理方面拥有 17 年以上的一手经验，横跨电商、网约车与开源生态。曾任步步高集团会员业务线产品经理，以及光彩神州集团的架构师与技术总监。目前担任翼飞开源负责人，并兼任 GitCoffee 开源社区的首席 Maintainer。 + +被评为 OpenTiny 开源社区（隶属华为云）2024 与 2025 年度优秀合作伙伴。深度聚焦于把开源社区建设与企业级 AI 应用打通。核心技术专长涵盖 Java 后端架构、全栈开发以及 AI Agent 落地实现。 \ No newline at end of file diff --git a/content/sessions/general-1215238.md b/content/sessions/general-1215238.md new file mode 100644 index 000000000..29613f1c0 --- /dev/null +++ b/content/sessions/general-1215238.md @@ -0,0 +1,19 @@ +--- +title: "Apache Seata's Growth Path and AI-driven Distributed Transaction Exploration" +date: "2026-08-08T15:45:00" +track: "general" +presenters: "Haoran Sun" +stype: "Chinese Session" +room: "Mtn BaiWang Hall" +--- + +This talk centers on Apache Seata's incubation and growth: the path from an Alibaba internal project to Apache incubation, community governance, ecosystem expansion, and release evolution. It analyzes key technical challenges and design decisions in multi-database, cross-language support, performance, and consistency guarantees. The second part presents AI-driven explorations for Seata, including ML-based anomaly transaction detection, intelligent conflict prediction and adaptive rollback strategies, automated performance bottleneck localization, and test-case generation — illustrated with practical case studies and outcomes. Suitable for audiences interested in distributed transaction implementations, open-source community-building, and AI+middleware practices. + +### Speakers: + + +
+ +Haoran Sun: Apache Seata Committer + +Apache Seata Committer, currently working on engine development in the Cloud-Native Storage Group at Xiaohongshu (RED). Junior student majoring in Software Engineering at Qingdao University of Science and Technology. Passionate about open source. \ No newline at end of file diff --git a/content/sessions/general-1215238.zh.md b/content/sessions/general-1215238.zh.md new file mode 100644 index 000000000..9421aa7c8 --- /dev/null +++ b/content/sessions/general-1215238.zh.md @@ -0,0 +1,19 @@ +--- +title: "Apache Seata 的成长之路与 AI 驱动的分布式事务探索" +date: "2026-08-08T15:45:00" +track: "general" +presenters: "孙浩然" +stype: "中文演讲" +room: "百望山会议室" +--- + +本次演讲围绕 Apache Seata 的孵化与成长展开：从阿里巴巴内部项目到进入 Apache 孵化，再到社区治理、生态扩展与版本演进的历程。我们将分析在多数据库、跨语言支持、性能和一致性保证等方面的关键技术挑战与设计决策。第二部分呈现 Seata 在 AI 驱动方向的探索，包括基于机器学习的异常事务检测、智能冲突预测与自适应回滚策略、自动化的性能瓶颈定位以及测试用例生成——并结合实际案例与成果加以说明。适合对分布式事务实现、开源社区建设以及 AI + 中间件实践感兴趣的听众。 + +### 讲师: + + +
+ +孙浩然：Apache Seata Committer + +Apache Seata Committer，目前在小红书云原生存储组做引擎开发，青岛科技大学软件工程大三学生，热爱开源 \ No newline at end of file diff --git a/content/sessions/incubator-1187157.md b/content/sessions/incubator-1187157.md new file mode 100644 index 000000000..1f3ecb566 --- /dev/null +++ b/content/sessions/incubator-1187157.md @@ -0,0 +1,25 @@ +--- +title: "Running an Apache Podling in the Real World: Lessons Beyond the Docs" +date: "2026-08-09T14:00:00" +track: "incubator" +presenters: "Shuxin Pan" +stype: "Chinese Session" +room: "Mtn YuQuan Hall" +--- + +The Apache Incubator is the gateway for projects joining the Apache Software Foundation, but for many new Podlings the real incubation journey involves much more than following official documentation. + +In practice, many important details—such as working effectively with mentors, bootstrapping Apache infrastructure, and navigating the first release vote—are often learned through experience rather than documentation. + +In this session, a PPMC member of Apache Fesod (Incubating) will share practical lessons learned from running a Podling in the real world. We will discuss how to evaluate whether a project is ready for the ASF, how to collaborate with mentors, how to bootstrap essential ASF infrastructure, and what really happens during a Podling’s first Apache release. + +Using Apache Fesod (Incubating) release practices as a case study, this session highlights common pitfalls and practical tips to help new Podlings navigate the Incubator and focus on building a sustainable open-source community aligned with The Apache Way. + +### Speakers: + + +
+ +Shuxin Pan: Apache Committer | Open Source Enthusiast + +Shuxin Pan is an open-source developer focused on Java distributed systems and AI. He serves as a PPMC member for Apache Fesod (Incubating) and a Committer for Apache Seata (Incubating) and Apache Ozhera (Incubating). With years of open-source contributions, He hopes to help more newcomers enter the world of open source. \ No newline at end of file diff --git a/content/sessions/incubator-1187157.zh.md b/content/sessions/incubator-1187157.zh.md new file mode 100644 index 000000000..d84dd9823 --- /dev/null +++ b/content/sessions/incubator-1187157.zh.md @@ -0,0 +1,25 @@ +--- +title: "在真实世界中运营一个 Apache 孵化项目：文档之外的实战经验" +date: "2026-08-09T14:00:00" +track: "incubator" +presenters: "Shuxin Pan" +stype: "中文演讲" +room: "玉泉山会议室" +--- + +Apache 孵化器（Incubator）是项目加入 Apache 软件基金会的入口，但对许多新晋 Podling（孵化项目）而言，真正的孵化之旅远不止照着官方文档按部就班。 + +在实践中，许多重要细节——如何与 mentors 高效协作、如何引导搭建 Apache 基础设施、如何走完第一次发布投票——往往要靠实战经验积累，而非文档所能涵盖。 + +在本次演讲中，一位 Apache Fesod（孵化中）的 PPMC 成员将分享在真实世界运营 Podling 的实战心得。我们将讨论如何评估一个项目是否已准备好进入 ASF、如何与 mentors 协作、如何搭建起关键的 ASF 基础设施，以及在一个 Podling 的首次 Apache 发布过程中到底会发生什么。 + +本次演讲将以 Apache Fesod（孵化中）的发布实践为案例，重点指出常见坑点与实用技巧，帮助新晋 Podling 顺利穿过孵化器，把精力集中在构建一个符合 The Apache Way、可持续发展的开源社区上。 + +### 讲师: + + +
+ +Shuxin Pan：Apache Committer | 开源爱好者 + +Shuxin Pan 是一名专注于 Java 分布式系统与 AI 的开源开发者。他是 Apache Fesod（孵化中）的 PPMC 成员，也是 Apache Seata（孵化中）和 Apache Ozhera（孵化中）的 Committer。凭借多年开源贡献经验，他希望能帮助更多新人走进开源世界。 \ No newline at end of file diff --git a/content/sessions/incubator-1210256.md b/content/sessions/incubator-1210256.md new file mode 100644 index 000000000..970d55fd6 --- /dev/null +++ b/content/sessions/incubator-1210256.md @@ -0,0 +1,27 @@ +--- +title: "Beyond Incubation: Building a Sustainable Apache Community with Cloudberry" +date: "2026-08-09T15:15:00" +track: "incubator" +presenters: "Dianjin Wang" +stype: "Chinese Session" +room: "Mtn YuQuan Hall" +--- + +Entering the Apache Incubator is only the beginning of an open-source project’s journey. The real challenge lies in what comes next: building a sustainable, diverse, and self-governing community that can thrive beyond its original creators. + +In this talk, we share practical lessons from Apache Cloudberry’s journey after incubation began—focusing on how we moved from a vendor-driven project to a community-driven ecosystem. Rather than revisiting the incubation process itself, this session dives into the often-overlooked challenges of scaling an Apache project in the real world. + +We will cover key aspects, including: +- Community scaling: growing beyond the initial team and fostering multi-organization contributions +- Governance in practice: applying the Apache Way in day-to-day decisions, from consensus building to handling disagreements +- Engineering and release discipline: evolving from ad-hoc processes to community-driven, repeatable release cycles +- Contributor growth: enabling contributors to become committers and future community leaders + +### Speakers: + + +
+ +Dianjin Wang: ALC Beijing Member, Track Chair of ApacheCon Asia 2021-2024, Head of Open Source at HashData + +ALC Beijing Member, Track Chair of ApacheCon Asia 2021-2024, Head of Open Source at HashData. \ No newline at end of file diff --git a/content/sessions/incubator-1210256.zh.md b/content/sessions/incubator-1210256.zh.md new file mode 100644 index 000000000..aa85273dc --- /dev/null +++ b/content/sessions/incubator-1210256.zh.md @@ -0,0 +1,27 @@ +--- +title: "孵化之后：用 Cloudberry 构建可持续的 Apache 社区" +date: "2026-08-09T15:15:00" +track: "incubator" +presenters: "Dianjin Wang" +stype: "中文演讲" +room: "玉泉山会议室" +--- + +进入 Apache 孵化器只是一个开源项目旅程的起点。真正的挑战在于接下来：构建一个可持续、多元、能够自我治理的社区，让它能在最初的创建者之外蓬勃发展。 + +在本次演讲中，我们将分享 Apache Cloudberry 自进入孵化以来的实战经验——重点讲述我们如何从一个厂商驱动的项目，转变为社区驱动的生态。本次演讲不会重述孵化流程本身，而是深入探讨在真实世界中扩展一个 Apache 项目时常被忽视的挑战。 + +我们将涵盖以下关键方面： +- 社区扩展：如何超越最初的核心团队、培育多组织贡献 +- 治理实践：在日常决策中践行 The Apache Way，从凝聚共识到处理分歧 +- 工程与发布纪律：从临时流程演进到社区驱动、可重复的发布周期 +- 贡献者成长：帮助贡献者成长为 Committer 与未来的社区领导者 + +### 讲师: + + +
+ +Dianjin Wang：ALC Beijing 成员，ApacheCon Asia 2021–2024 议题主席，HashData 开源负责人 + +ALC Beijing 成员，ApacheCon Asia 2021–2024 议题主席，HashData 开源负责人。 \ No newline at end of file diff --git a/content/sessions/incubator-1210342.md b/content/sessions/incubator-1210342.md new file mode 100644 index 000000000..7d5a7ffee --- /dev/null +++ b/content/sessions/incubator-1210342.md @@ -0,0 +1,28 @@ +--- +title: "GeaFlow｜Streaming Graph Computing Engine and AI Memory Practice" +date: "2026-08-09T13:30:00" +track: "incubator" +presenters: "Litao Lin, Yao Zhongqiang" +stype: "Chinese Session" +room: "Mtn YuQuan Hall" +--- + +GeaFlow is currently incubating at the Apache Software Foundation (ASF). As the industry's first open-source engine dedicated to streaming graph computing, it addresses the challenges that traditional static graph computing faces in dynamic scenarios amid the explosion of real-time data. Through unified semantic design and scheduling models for streaming-batch graph processing, GeaFlow enables efficient analysis and computation of trillion-scale graphs, redefining the technical boundaries of graph computing. +This talk will explore the intrinsic connection between streaming graph computing and real-time Agent memory. We will demonstrate how GeaFlow leverages the inherent capabilities of incremental computation and state management in streaming graphs to solve the real-time update and consistency challenges faced by Agent memory. We will elaborate on the design of GraphMemory, exploring how a hybrid storage architecture that combines vectors, text, and graph relationships can address the limitations of memory retrieval and enhance the long-term memory recall and reasoning capabilities of Agents. +In the technical practice section, we will analyze the application of Graph-GNN (such as SAGNN) in vertical domains like intelligent transportation. We will demonstrate how to construct dynamic road network graphs using GeaFlow and achieve real-time write-back and updates of graph states based on road segment impedance predictions. Additionally, we will outline the architectural roadmap for GeaFlow's integration with other open source engine systems to support trillion-scale graph computing. + +### Speakers: + + +
+ +Litao Lin: Apache GeaFlow (Incubating) Committer + +Apache GeaFlow (Incubating) Committer. As a core member of the project, he participated in the architecture design and development of the GeaFlow graph computing engine from scratch, with a particular focus on the design and implementation of the graph computing DSL and the evolution of data intelligence technologies. He is deeply involved in open source community activities and currently works at Ant Group. + + +
+ +Yao Zhongqiang: Graph Computing Expert & Development Engineer at Ant Group + +Deeply involved in the big data field, specializing in graph computing, real-time computing, and OLAP. Current team focus is on Agent Memory and MARL. \ No newline at end of file diff --git a/content/sessions/incubator-1210342.zh.md b/content/sessions/incubator-1210342.zh.md new file mode 100644 index 000000000..599c9d7b8 --- /dev/null +++ b/content/sessions/incubator-1210342.zh.md @@ -0,0 +1,28 @@ +--- +title: "GeaFlow｜流式图计算引擎与 AI 记忆实践" +date: "2026-08-09T13:30:00" +track: "incubator" +presenters: "Litao Lin, Yao Zhongqiang" +stype: "中文演讲" +room: "玉泉山会议室" +--- + +GeaFlow 目前正在 Apache 软件基金会（ASF）孵化。作为业界首个专注于流式图计算的开源引擎，它应对的是实时数据爆发背景下，传统静态图计算在动态场景中所面临的挑战。通过统一的语义设计和面向流批一体的图处理调度模型，GeaFlow 能够对万亿规模的图进行高效分析与计算，重新定义了图计算的技术边界。 +本次演讲将探讨流式图计算与实时 Agent 记忆之间的内在联系。我们将演示 GeaFlow 如何借助流式图在增量计算和状态管理上的天然能力，解决 Agent 记忆面临的实时更新与一致性难题。我们将深入阐述 GraphMemory 的设计，探讨融合向量、文本与图关系的混合存储架构如何突破记忆检索的局限，增强 Agent 的长期记忆召回与推理能力。 +在技术实践部分，我们将分析 Graph-GNN（如 SAGNN）在智能交通等垂直领域的应用，演示如何用 GeaFlow 构建动态路网图，并基于路段阻抗预测实现图状态的实时写回与更新。此外，我们还会勾勒 GeaFlow 与其他开源引擎系统集成的架构路线图，以支撑万亿级图计算。 + +### 讲师: + + +
+ +Litao Lin：Apache GeaFlow（孵化中）Committer + +Apache GeaFlow（孵化中）Committer。作为项目核心成员，他从零参与 GeaFlow 图计算引擎的架构设计与开发，尤其专注于图计算 DSL 的设计与实现以及数据智能技术的演进。他深度参与开源社区活动，目前就职于蚂蚁集团。 + + +
+ +Yao Zhongqiang：蚂蚁集团图计算专家与开发工程师 + +深耕大数据领域，专注于图计算、实时计算与 OLAP。当前团队工作重心在 Agent Memory 与多智能体强化学习（MARL）。 \ No newline at end of file diff --git a/content/sessions/incubator-1212002.md b/content/sessions/incubator-1212002.md new file mode 100644 index 000000000..5fd8812f4 --- /dev/null +++ b/content/sessions/incubator-1212002.md @@ -0,0 +1,34 @@ +--- +title: "Build Once, Run on Any Linux: SynxDB CE for Apache Cloudberry (Incubating)" +date: "2026-08-09T15:45:00" +track: "incubator" +presenters: "Shine Zhang" +stype: "English Session" +room: "Mtn YuQuan Hall" +--- + +Apache Cloudberry (Incubating) is a Postgres-compatible MPP analytical database with a fast-growing community. Like many Apache projects, its official releases are source tarballs — exactly right for ASF governance, but leaving a gap for the people who actually want to run it. Hobbyists, evaluators, and enterprise teams all expect an installable binary. + +The obvious approach — "just build one package per distro" — is a trap. glibc symbol versions, compiler ABIs, bundled library versions (OpenSSL, Perl, Python), and RPM-vs-DEB packaging all multiply the matrix until you're maintaining a dozen variants of the same release. + +For SynxDB CE, our Community Edition of Apache Cloudberry (Incubating), we took a different path: build once, run on any Linux where glibc is recent enough. The strategy: + +Bundle the entire toolchain. A layered Docker build environment (toolchain → devel-base → devel-cbdb → runtime) ships our own GCC 12.2.1, binutils, Perl, Python, cmake, Conan, meson, and ninja — all built from source into /usr/local/toolchain/. The host OS contributes nothing to the compile. Every build, on every developer's laptop and every CI runner, uses the same bits. + +Target glibc 2.17 as the ABI floor, using CentOS 7 as the build platform. glibc's backward-compatibility guarantee means a binary linked against 2.17 runs unchanged on Rocky 8 (glibc 2.28), Rocky 9 (2.34), Ubuntu 22.04 (2.35), Ubuntu 24.04 (2.39), and SUSE releases of similar vintage. + +Make glibc the only runtime dependency. Everything else — OpenSSL, libhdfs3, ORC, Parquet, libgsasl, libftp, liboss2 — is vendored through a Conan 2.x dependency DAG and linked into the distribution. No external package-manager dependencies. No "install these 40 RPMs first." +Package from one binary tree into both RPM and DEB. Single build, two artifacts. No parallel pipelines. + +Verify portability as a CI gate. ldd scans confirm no unexpected shared-library edges; smoke tests install the DEB on Ubuntu 24.04 and run SELECT version(). + +We'll walk through the Dockerfile hierarchy, show the Conan package DAG, and talk candidly about what didn't work on the way here — compiler ABI pitfalls, glibc symbol surprises, and the quiet cost of fpm. + +### Speakers: + + +
+ +Shine Zhang: CTO and co-founder of Synx Data Labs + +Xin Zhang (Shine) — CTO and co-founder of Synx Data Labs. Former Pivotal/VMware Greenplum engineer, PostgreSQL contributor, and long-time participant in the Greenplum / Cloudberry lineage. Leads Synx's upstream contributions to Apache Cloudberry (Incubating) and the SynxDB Community Edition release pipeline. \ No newline at end of file diff --git a/content/sessions/incubator-1212002.zh.md b/content/sessions/incubator-1212002.zh.md new file mode 100644 index 000000000..311d68f8c --- /dev/null +++ b/content/sessions/incubator-1212002.zh.md @@ -0,0 +1,35 @@ +--- +title: "一次构建，任意 Linux 运行：面向 Apache Cloudberry（孵化中）的 SynxDB CE" +date: "2026-08-09T15:45:00" +track: "incubator" +presenters: "Shine Zhang" +stype: "英文演讲" +room: "玉泉山会议室" +--- + +Apache Cloudberry（孵化中）是一个兼容 Postgres 的 MPP 分析型数据库，社区增长迅速。与许多 Apache 项目一样，它的官方发布物是源码 tarball——这完全符合 ASF 的治理要求，却给真正想运行它的人留下了一段空白。爱好者、评估者以及企业团队，都期望能有一个可直接安装的二进制包。 + +最显而易见的做法——"每个发行版各打一个包"——其实是个陷阱。glibc 符号版本、编译器 ABI、内置库版本（OpenSSL、Perl、Python），再加上 RPM 与 DEB 打包方式的差异，会让构建矩阵不断膨胀，直到你为同一个版本维护十几种变体。 + +对于 SynxDB CE——我们的 Apache Cloudberry（孵化中）社区版——我们走了一条不同的路：一次构建，在任何 glibc 版本足够新的 Linux 上运行。策略如下： + +打包整套工具链。一个分层的 Docker 构建环境（toolchain → devel-base → devel-cbdb → runtime）搭载了我们自带的 GCC 12.2.1、binutils、Perl、Python、cmake、Conan、meson 和 ninja——全部从源码构建进 /usr/local/toolchain/。宿主 OS 对编译过程不贡献任何东西。每一次构建，无论在哪个开发者的笔记本上、哪一台 CI runner 上，用的都是同一套二进制。 + +以 glibc 2.17 作为 ABI 下限，并使用 CentOS 7 作为构建平台。glibc 向后兼容的承诺意味着：链接于 2.17 的二进制，可以原封不动地运行在 Rocky 8（glibc 2.28）、Rocky 9（2.34）、Ubuntu 22.04（2.35）、Ubuntu 24.04（2.39），以及年代相近的 SUSE 发行版上。 + +让 glibc 成为唯一的运行时依赖。其余一切——OpenSSL、libhdfs3、ORC、Parquet、libgsasl、libftp、liboss2——都通过 Conan 2.x 的依赖 DAG 内嵌进来并链接进发行包。没有对包管理器的外部依赖。再也不需要"先装这 40 个 RPM"。 + +从同一棵二进制树打包出 RPM 和 DEB。一次构建，两份产物。无需并行流水线。 + +把可移植性校验作为一道 CI 关卡。ldd 扫描确认没有意外的共享库依赖边；冒烟测试在 Ubuntu 24.04 上安装 DEB 并运行 SELECT version()。 + +我们将走查这套 Dockerfile 层级结构，展示 Conan 包的 DAG，并坦诚地聊聊一路上没走通的方案——编译器 ABI 的坑、glibc 符号的意外，以及 fpm 悄悄带来的代价。 + +### 讲师: + + +
+ +Shine Zhang：Synx Data Labs 联合创始人兼 CTO + +Xin Zhang（Shine）——Synx Data Labs 联合创始人兼 CTO。前 Pivotal/VMware Greenplum 工程师、PostgreSQL 贡献者，长期参与 Greenplum / Cloudberry 这一技术脉络。负责 Synx 对 Apache Cloudberry（孵化中）的上游贡献，以及 SynxDB 社区版的发布流水线。 \ No newline at end of file diff --git a/content/sessions/incubator-1213722.md b/content/sessions/incubator-1213722.md new file mode 100644 index 000000000..8484016c5 --- /dev/null +++ b/content/sessions/incubator-1213722.md @@ -0,0 +1,25 @@ +--- +title: "AI-assisted Apache Incubator oversight" +date: "2026-08-09T16:15:00" +track: "incubator" +presenters: "Justin Mclean" +stype: "English Session" +room: "Mtn YuQuan Hall" +--- + +Managing multiple podlings across the Apache Incubator requires synthesising data from multiple disconnected sources: Confluence reports, GitHub activity, mailing lists, Whimsy, with no single view of the full picture. This talk introduces a set of Model Context Protocol (MCP) servers built to change that. +The tooling spans multiple connected servers: a podlings server reading lifecycle data directly from podlings.xml, an apache-health server exposing commit and community metrics, an incubator-mail server providing searchable access to general@incubator.apache.org, and an IPMC synthesis layer that combines all three into watchlists, graduation readiness assessments, podling briefs, and stall detection. +The talk will demonstrate real workflows: monthly IPMC oversight passes, mentor check-ins on specific podlings, identifying chronically stalled projects, and cross-referencing report narratives against health metrics to spot where the two tell different stories. It will also explore a bigger question the tooling raises - Could AI-assisted oversight change the need, frequency, or format of podling reports themselves? + +### Speakers: + + +
+ +Justin Mclean: ASF Director, VP ASF Incubator, Datastrato Community Manager + +Justin Mclean is a long-time contributor to the Apache Software Foundation (ASF) and currently serves as VP of the ASF Incubator and a member of the ASF Board of Directors. Over several years, he has contributed to evolving the Incubator’s governance, policies, and training programs, including documenting how projects operate in practice and analysing the Incubator’s history. Through this work, he has supported hundreds of open source projects in building sustainable and independent communities. + +He has extensive experience guiding projects from early incubation through to successful graduation, and has reviewed over a thousand open source releases. His work focuses on open source governance, community health, and the design of systems that enable collaboration at scale. + +Justin regularly speaks at international conferences on topics such as open source governance, licensing, and community building, sharing practical lessons drawn from real-world experience across a wide range of ASF projects and communities. \ No newline at end of file diff --git a/content/sessions/incubator-1213722.zh.md b/content/sessions/incubator-1213722.zh.md new file mode 100644 index 000000000..ac8343e84 --- /dev/null +++ b/content/sessions/incubator-1213722.zh.md @@ -0,0 +1,27 @@ +--- +title: "AI 辅助的 Apache 孵化器监督" +date: "2026-08-09T16:15:00" +track: "incubator" +presenters: "Justin Mclean" +stype: "英文演讲" +room: "玉泉山会议室" +--- + +在 Apache 孵化器中同时管理多个 Podling，需要整合来自多个互不相连来源的数据：Confluence 报告、GitHub 活动、邮件列表、Whimsy，却没有任何一个地方能呈现完整全貌。本次演讲将介绍一组为改变这一现状而构建的模型上下文协议（MCP）服务器。 + +这套工具横跨多个相互连接的服务器：一个直接从 podlings.xml 读取生命周期数据的 podlings 服务器，一个暴露提交与社区指标的 apache-health 服务器，一个提供 general@incubator.apache.org 可检索访问的 incubator-mail 服务器，以及一个 IPMC 综合层——它把三者合并成关注列表、毕业就绪度评估、Podling 简报和停滞检测。 + +演讲将演示真实的工作流：每月一次的 IPMC 监督通查、针对特定 Podling 的 mentor 跟进、识别长期停滞的项目，以及把报告叙述与健康指标交叉比对，找出两者说法不一致之处。它还会探讨这套工具引出的一个更大问题——AI 辅助的监督，是否会改变 Podling 报告本身的需求、频率或形式？ + +### 讲师: + + +
+ +Justin Mclean：ASF 董事、ASF 孵化器 VP、Datastrato 社区经理 + +Justin Mclean 是 Apache 软件基金会（ASF）的长期贡献者，目前担任 ASF 孵化器 VP 及 ASF 董事会成员。多年来，他致力于推动孵化器的治理、政策与培训项目的演进，包括把项目实际运作方式整理成文，以及对孵化器历史进行分析。通过这些工作，他已帮助数以百计的开源项目构建起可持续、独立的社区。 + +在引导项目从早期孵化走向成功毕业方面，他经验丰富，审阅过的开源发布物已逾千个。他的工作聚焦于开源治理、社区健康，以及支撑大规模协作的系统设计。 + +Justin 经常在国际会议上就开源治理、许可证与社区建设等主题发表演讲，分享来自各类 ASF 项目与社区真实实践的鲜活经验。 \ No newline at end of file diff --git a/content/sessions/incubator-1214505.md b/content/sessions/incubator-1214505.md new file mode 100644 index 000000000..5c51402f2 --- /dev/null +++ b/content/sessions/incubator-1214505.md @@ -0,0 +1,29 @@ +--- +title: "Apache GraphAr (Incubating): Graph Data Management Format for the AI Era" +date: "2026-08-09T14:30:00" +track: "incubator" +presenters: "Xiaokang Yang, Zhe Wang" +stype: "Chinese Session" +room: "Mtn YuQuan Hall" +--- + +With the rapid evolution of AI, the application scenarios for graph data have become remarkably diverse and widespread. Iterative graph computing, Graph Neural Networks (GNN), and GraphRAG are driving growing demand for graph-structured data across industries. However, in most enterprises today, graph data remains siloed -- hidden in log files, scattered across MySQL databases, or locked inside proprietary graph databases. When different scenarios require graph data to flow between systems, engineers face repeated ETL efforts, format incompatibilities, and significant friction. Meanwhile, a fundamental question arises: should we always load graph data into a transactional graph database just to serve diverse workloads? Doing so often leads to unnecessary data redundancy and operational overhead. + +Fortunately, Apache GraphAr (Incubating) effectively addresses this challenge by providing an open graph storage format.In this talk, we will introduce GraphAr's core philosophy of enabling seamless graph data flow across systems, and share recent collaborations with different open-source communities that put this vision into practice. We will also discuss the community operations of GraphAr within the Apache Incubator -- how we build a diverse contributor base, practice the Apache Way, and plan the roadmap toward graduation. This session is designed for developers, data engineers, and open-source enthusiasts interested in graph technologies, data lake architecture, or community building under the Apache umbrella. + + +### Speakers: + + +
+ +Xiaokang Yang: Apache GraphAr (Incubating) Committer + +Xiaokang Yang is a Committer of Apache GraphAr (Incubating), an open-source project that provides a standardized file format and multi-language libraries for graph data storage in data lakes. He is currently a Master's student at Northeastern University (China) and an intern at Alibaba, with a research focus on graph computing. He has authored two research papers in the graph domain. He is passionate about graph data management and building open-source communities following the Apache Way. + + +
+ +Zhe Wang: Apache GraphAr(incubating) PPMC + +Master student at JLU, interested in graph store and graph AI. \ No newline at end of file diff --git a/content/sessions/incubator-1214505.zh.md b/content/sessions/incubator-1214505.zh.md new file mode 100644 index 000000000..ae977ad20 --- /dev/null +++ b/content/sessions/incubator-1214505.zh.md @@ -0,0 +1,29 @@ +--- +title: "Apache GraphAr（孵化中）：面向 AI 时代的图数据管理格式" +date: "2026-08-09T14:30:00" +track: "incubator" +presenters: "Xiaokang Yang, Zhe Wang" +stype: "中文演讲" +room: "玉泉山会议室" +--- + +随着 AI 的快速演进，图数据的应用场景变得异常多元和广泛。迭代式图计算、图神经网络（GNN）和 GraphRAG 正在推动各行各业对图结构数据的需求不断增长。然而，在当今大多数企业中，图数据仍处于孤岛状态——藏在日志文件里，散落在 MySQL 数据库中，或被锁在专有图数据库之内。当不同场景需要图数据在系统间流转时，工程师们要反复应对 ETL 工作量、格式不兼容，以及显著的摩擦。与此同时，一个根本性的问题随之而来：我们是否总是要把图数据加载到事务型图数据库中，才能服务于各种工作负载？这样做往往带来不必要的数据冗余与运维开销。 + +幸运的是，Apache GraphAr（孵化中）通过提供一种开放的图存储格式，有效地应对了这一挑战。在本次演讲中，我们将介绍 GraphAr 让图数据在系统间无缝流动的核心理念，并分享我们与不同开源社区近期合作、把这一愿景付诸实践的案例。我们还会讨论 GraphAr 在 Apache 孵化器内的社区运营——我们如何构建多元化的贡献者基础、践行 The Apache Way，以及规划走向毕业的路线图。本次演讲面向对图技术、数据湖架构，或在 Apache 体系下进行社区建设感兴趣的开发者、数据工程师与开源爱好者。 + + +### 讲师: + + +
+ +Xiaokang Yang：Apache GraphAr（孵化中）Committer + +Xiaokang Yang 是 Apache GraphAr（孵化中）的 Committer——这是一个为数据湖中的图数据存储提供标准化文件格式与多语言库的开源项目。他目前是东北大学（中国）的硕士研究生，在阿里巴巴实习，研究方向为图计算。他在图领域发表过两篇研究论文。他热衷于图数据管理，并致力于遵循 The Apache Way 建设开源社区。 + + +
+ +Zhe Wang：Apache GraphAr（孵化中）PPMC + +吉林大学硕士研究生，兴趣方向为图存储与图 AI。 \ No newline at end of file diff --git a/content/sessions/iot-1185186.md b/content/sessions/iot-1185186.md new file mode 100644 index 000000000..9ab8ff201 --- /dev/null +++ b/content/sessions/iot-1185186.md @@ -0,0 +1,19 @@ +--- +title: "Apache NIFI , what been updated since 2025" +date: "2026-08-09T15:15:00" +track: "iot" +presenters: "Yan Liu" +stype: "English Session" +room: "Mtn BaiWang Hall" +--- + +I will introduce the recent improvements in Apache NIFI since 2025, including new features and how enterprise use Apache NIFI to help their daily works + +### Speakers: + + +
+ +Yan Liu: Cloudera + +Cloudera Solution Engineering . Apache Hive and Apache Flink Contributor. Over 10 Years of Practical Experience in Big Data and my current focus is real-time data warehouse using Apache Flink, Apache Hive, and Apache Iceberg. \ No newline at end of file diff --git a/content/sessions/iot-1185186.zh.md b/content/sessions/iot-1185186.zh.md new file mode 100644 index 000000000..843472064 --- /dev/null +++ b/content/sessions/iot-1185186.zh.md @@ -0,0 +1,19 @@ +--- +title: "Apache NiFi：2025 年以来的更新" +date: "2026-08-09T15:15:00" +track: "iot" +presenters: "Yan Liu" +stype: "英文演讲" +room: "百望山会议室" +--- + +我将介绍 Apache NiFi 自 2025 年以来的最新改进，包括新特性以及企业如何利用 Apache NiFi 辅助其日常工作 + +### 讲师: + + +
+ +Yan Liu: Cloudera + +Cloudera 解决方案工程部。Apache Hive 和 Apache Flink 贡献者。在大数据领域拥有超过 10 年的实践经验，目前专注于使用 Apache Flink、Apache Hive 和 Apache Iceberg 构建实时数据仓库。 \ No newline at end of file diff --git a/content/sessions/iot-1196529.md b/content/sessions/iot-1196529.md new file mode 100644 index 000000000..b1f55322d --- /dev/null +++ b/content/sessions/iot-1196529.md @@ -0,0 +1,39 @@ +--- +title: "When MQTT Meets Kafka: Evolving IoT Infrastructure from Ingestion Layer to Data Layer" +date: "2026-08-09T14:30:00" +track: "iot" +presenters: "Lobo Xu" +stype: "Chinese Session" +room: "Mtn BaiWang Hall" +--- + +Data generated by IoT devices is being consumed by an ever-growing number of downstream systems — real-time analytics, model training, AI agent decision-making. The familiar engineering playbook is to let MQTT handle device connectivity and Kafka handle the data pipeline, gluing the two together with a bridge layer. This architecture works well enough in traditional IoT scenarios. But when the number of downstream consumers grows from one to five, when the same data needs to feed both a training pipeline and a real-time inference service simultaneously, the cracks in the bridge layer become hard to ignore: unpredictable latency, silent data loss, broken backpressure propagation. + +The root cause is not the bridge layer itself, but the positioning of the MQTT Broker — designed as an ingestion layer, eager to push data out the moment it arrives, with no built-in durability or stream consumption capabilities. Every new downstream use case means another bridge link and another operational blind spot. + +This talk shares architectural lessons from RobustMQ, an open-source unified message broker built in Rust. The core idea: a unified storage layer that natively supports both MQTT and Kafka protocols — data written by devices over MQTT can be consumed directly via the Kafka protocol, with no bridging, no data copying, one message, one copy, consumed by any protocol. + +We will share: + • Real pain points of MQTT + Kafka bridge architectures in multi-consumer IoT scenarios + • Key architectural decisions in evolving the MQTT Broker from "ingestion layer" to "data layer" + • The unified storage layer design: one copy of data, consumed natively by multiple protocols + • The full edge-to-cloud pipeline: offline buffering, auto-sync, minimal deployment footprint + • Current progress, known limitations, and future direction + +This is an ongoing architectural exploration, not a success story. RobustMQ is fully open-source under the Apache 2.0 License. We look forward to exchanging real-world experience with engineers who are tackling the same IoT data plumbing problems, and we hope to eventually become an Apache Software Foundation top-level project, giving back to the open-source community. + + +https://robustmq.com/ + +https://github.com/robustmq/robustmq + +### Speakers: + + +
+ +Lobo Xu: RobustMQ Auther + +Author of RobustMQ, former senior engineer of ByteDance, former senior engineer of Tencent Cloud. Former Head of Kafka/RabbitMQ technology at Tencent Cloud message Queue. +Author of RobustMQ and Apache Kafka/RocketMQ/Pulsar Contributor is active in the message queuing and Serverless open source communities. Deep product, technical and commercial understanding and vision of MQ and Serverless. +He is the author of the Geek Hour columns "Breaking Down Message Queues in Depth" and "Rust in Action: Writing the Next Generation of Cloud-Native Message Queues" and co-author of the technical book "The Road to Foundational Software: Enterprise Practices and Open Source Strategies." \ No newline at end of file diff --git a/content/sessions/iot-1196529.zh.md b/content/sessions/iot-1196529.zh.md new file mode 100644 index 000000000..19171a091 --- /dev/null +++ b/content/sessions/iot-1196529.zh.md @@ -0,0 +1,39 @@ +--- +title: "当 MQTT 遇见 Kafka：IoT 基础设施从数据接入层到数据层的演进" +date: "2026-08-09T14:30:00" +track: "iot" +presenters: "Lobo Xu" +stype: "中文演讲" +room: "百望山会议室" +--- + +IoT 设备产生的数据正被越来越多的下游系统消费——实时分析、模型训练、AI 智能体决策。大家熟悉的工程套路是：让 MQTT 负责设备连接，Kafka 负责数据管道，再用一个桥接层把两者粘合起来。这种架构在传统 IoT 场景下表现尚可。但当下游消费者从一个增加到五个，当同一份数据需要同时喂给训练管道和实时推理服务时，桥接层的裂缝就难以忽视了：不可预测的延迟、悄无声息的数据丢失、断裂的反压传导。 + +根因并不在桥接层本身，而在于 MQTT Broker 的定位——它被设计为数据接入层，数据一到就急于推送出去，没有内置的持久化或流式消费能力。每新增一个下游用例，就意味着多一条桥接链路、多一个运维盲区。 + +本次演讲将分享来自 RobustMQ 的架构经验——这是一个用 Rust 构建的开源统一消息中间件。其核心理念是：一个原生同时支持 MQTT 和 Kafka 协议的统一存储层——设备通过 MQTT 写入的数据可以直接通过 Kafka 协议消费，无需桥接、无需数据复制，一条消息、一份拷贝，可被任意协议消费。 + +我们将分享： + • 多消费者 IoT 场景下 MQTT + Kafka 桥接架构的真实痛点 + • 将 MQTT Broker 从"数据接入层"演进为"数据层"的关键架构决策 + • 统一存储层设计：一份数据，由多种协议原生消费 + • 从边缘到云的完整管道：离线缓冲、自动同步、最小化部署 footprint + • 当前进度、已知局限与未来方向 + +这是一次正在进行的架构探索，而非一个成功故事。RobustMQ 在 Apache 2.0 许可证下完全开源。我们期待与同样在解决 IoT 数据管道问题的工程师交流实战经验，也希望最终能成为 Apache 软件基金会的顶级项目，回馈开源社区。 + + +https://robustmq.com/ + +https://github.com/robustmq/robustmq + +### 讲师: + + +
+ +Lobo Xu: RobustMQ Auther + +RobustMQ 作者，前字节跳动高级工程师，前腾讯云高级工程师。曾任腾讯云消息队列 Kafka/RabbitMQ 技术负责人。 +作为 RobustMQ 作者及 Apache Kafka/RocketMQ/Pulsar 贡献者，他活跃于消息队列和 Serverless 开源社区，对 MQ 和 Serverless 有着深入的产品、技术与商业理解及前瞻视野。 +他是极客时间专栏《深入拆解消息队列》和《Rust 实战：编写下一代云原生消息队列》的作者，并合著了技术书籍《基础软件之路：企业实践与开源策略》。 \ No newline at end of file diff --git a/content/sessions/iot-1213290.md b/content/sessions/iot-1213290.md new file mode 100644 index 000000000..6d00743bc --- /dev/null +++ b/content/sessions/iot-1213290.md @@ -0,0 +1,25 @@ +--- +title: "From Time Series Data to Specialized Data Management" +date: "2026-08-09T15:45:00" +track: "iot" +presenters: "洪胤 张, 泽宇 张" +stype: "Chinese Session" +room: "Mtn BaiWang Hall" +--- + +This session focuses on the evolution of data management from time-series data to specialized data. As IoT, scientific computing, and intelligent systems continue to grow, data management is expanding from classic time-series workloads to more complex data types, including spatiotemporal, gridded, remote sensing, meteorological, log, and multimodal data. The session will discuss the new challenges these data bring to storage, indexing, querying, and unified access, and explore emerging approaches such as format-aware queries, cross-file organization, and database–file system integration. + +### Speakers: + + +
+ +洪胤 张: Apache IoTDB PMC, PhD student at Tsinghua University + +Zhang Hongyin, Apache IoTDB PMC, PhD student of Tsinghua University, is currently responsible for Apache IoTDB memory management, system monitoring, and standardized testing. + +
+ +泽宇 张: Graduate student at Tsinghua University + +Zhang Zeyu, Graduate student of Tsinghua University, is currently focusing on Apache IoTDB agent-friendly system, along with time series model training and reasoning. \ No newline at end of file diff --git a/content/sessions/iot-1213290.zh.md b/content/sessions/iot-1213290.zh.md new file mode 100644 index 000000000..137107b57 --- /dev/null +++ b/content/sessions/iot-1213290.zh.md @@ -0,0 +1,25 @@ +--- +title: "从时序数据到专业化数据管理" +date: "2026-08-09T15:45:00" +track: "iot" +presenters: "洪胤 张, 泽宇 张" +stype: "中文演讲" +room: "百望山会议室" +--- + +本次演讲聚焦数据管理从时序数据到专业化数据的演进。随着 IoT、科学计算和智能系统的持续发展，数据管理正在从经典的时序工作负载扩展到更复杂的数据类型，包括时空数据、网格数据、遥感数据、气象数据、日志数据和多模态数据。本次演讲将讨论这些数据给存储、索引、查询和统一访问带来的新挑战，并探索格式感知查询、跨文件组织以及数据库与文件系统融合等新兴方法。 + +### 讲师: + + +
+ +洪胤 张: Apache IoTDB PMC, PhD student at Tsinghua University + +张洪胤，Apache IoTDB PMC，清华大学博士研究生，目前负责 Apache IoTDB 的内存管理、系统监控和标准化测试工作。 + +
+ +泽宇 张: Graduate student at Tsinghua University + +张泽宇，清华大学本科毕业生，目前聚焦 Apache IoTDB agent-friendly system，以及时间序列模型训练和推理。 \ No newline at end of file diff --git a/content/sessions/iot-1213309.md b/content/sessions/iot-1213309.md new file mode 100644 index 000000000..75c683f1b --- /dev/null +++ b/content/sessions/iot-1213309.md @@ -0,0 +1,23 @@ +--- +title: "openvela AI Agent: Building the Next-Generation Proactive Intelligent Hardware Ecosystem" +date: "2026-08-09T13:30:00" +track: "iot" +presenters: "Xingyu Yan" +stype: "Chinese Session" +room: "Mtn BaiWang Hall" +--- + +AI Agents are moving from the cloud to edge devices. openvela is an open-source operating system built on Apache NuttX, deployed on over 100 million devices. This talk introduces the openvela lightweight native AI Agent engine — a pure C implementation requiring as little as 100KB RAM — enabling watches, speakers, glasses, earbuds, and other AI hardware to make autonomous decisions and deliver proactive services.Outline: +1. The Trend: AI Agents Moving from Cloud to Edge +2. openvela AI Agent Architecture Deep Dive +3. Real-World Scenario Demonstrations +4. Open Source Ecosystem and Future Outlook + +### Speakers: + + +
+ +Xingyu Yan: openvela Technical Evangelist + +Xingyu Yan is an openvela Technical Evangelist with experience in embedded systems development. He is responsible for the openvela hardware ecosystem, including development board enablement and expansion. He is also actively involved in promoting the openvela AI Agent engine across smart hardware platforms. \ No newline at end of file diff --git a/content/sessions/iot-1213309.zh.md b/content/sessions/iot-1213309.zh.md new file mode 100644 index 000000000..69ac4f5ba --- /dev/null +++ b/content/sessions/iot-1213309.zh.md @@ -0,0 +1,23 @@ +--- +title: "openvela AI Agent：构建下一代主动式智能硬件生态" +date: "2026-08-09T13:30:00" +track: "iot" +presenters: "Xingyu Yan" +stype: "中文演讲" +room: "百望山会议室" +--- + +AI 智能体正在从云端走向边缘设备。openvela 是一个基于 Apache NuttX 构建的开源操作系统，已部署在超过 1 亿台设备上。本次演讲将介绍 openvela 轻量级原生 AI Agent 引擎——一个纯 C 实现，仅需 100KB 内存——使手表、音箱、眼镜、耳机等各类 AI 硬件能够自主决策并提供主动式服务。大纲： +1. 趋势：AI 智能体从云端走向边缘 +2. openvela AI Agent 架构深度解析 +3. 真实场景演示 +4. 开源生态与未来展望 + +### 讲师: + + +
+ +Xingyu Yan: openvela Technical Evangelist + +Xingyu Yan 是 openvela 技术布道师，拥有嵌入式系统开发经验。他负责 openvela 硬件生态，包括开发板的适配与扩展。他还积极推动 openvela AI Agent 引擎在各类智能硬件平台上的落地。 \ No newline at end of file diff --git a/content/sessions/iot-1213890.md b/content/sessions/iot-1213890.md new file mode 100644 index 000000000..f5a6b2a0e --- /dev/null +++ b/content/sessions/iot-1213890.md @@ -0,0 +1,19 @@ +--- +title: "Unifying Robotic Data: Managing ROS Messages with Apache TsFile on the Edge" +date: "2026-08-09T14:00:00" +track: "iot" +presenters: "Ansi Zhang" +stype: "Chinese Session" +room: "Mtn BaiWang Hall" +--- + +The Robot Operating System (ROS) is the de facto standard for robotics development, yet efficiently managing the vast amounts of time-series data generated on edge devices remains a challenge. This session introduces a new open-source tool that bridges the ROS ecosystem with the Apache big data world. We will demonstrate how to convert ROS messages directly into Apache TsFile, a high-performance, columnar file format designed for IoT time-series data. This approach enables efficient local storage, compression, and seamless data synchronization with Apache IoTDB in the cloud. Attendees will learn about the tool's architecture, see a live demo of edge data collection from a simulated robot, and understand how this pattern can be extended to build a complete edge-to-cloud data pipeline for robotics and industrial IoT. + +### Speakers: + + +
+ +Ansi Zhang: ROS + Apache TsFile | Edge Data Management + +Ansi Zhang is an Associate Professor at Guizhou University, as well as a Visiting Scholar at Tsinghua University. His research focuses on UAV Big Data and Industrial Big Data, areas where efficient edge data management is critical. He has led 7 national and provincial research projects, published over 20 SCI/EI papers, and holds 19 patents and 10 software copyrights. As the creator of the ROS-to-TsFile tool, he bridges the gap between robotic data collection and the Apache big data ecosystem, enabling lightweight, high-compression storage for autonomous systems at the edge. At Community Over Code, he looks forward to sharing how Apache TsFile can revolutionize data pipelines in robotics and IIoT. \ No newline at end of file diff --git a/content/sessions/iot-1213890.zh.md b/content/sessions/iot-1213890.zh.md new file mode 100644 index 000000000..68140626f --- /dev/null +++ b/content/sessions/iot-1213890.zh.md @@ -0,0 +1,19 @@ +--- +title: "统一机器人数据：在边缘侧用 Apache TsFile 管理 ROS 消息" +date: "2026-08-09T14:00:00" +track: "iot" +presenters: "Ansi Zhang" +stype: "中文演讲" +room: "百望山会议室" +--- + +机器人操作系统（ROS）是机器人开发领域事实上的标准，但如何高效管理边缘设备上产生的大量时序数据仍是一项挑战。本次演讲将介绍一款新的开源工具，它将 ROS 生态与 Apache 大数据世界连接起来。我们将演示如何将 ROS 消息直接转换为 Apache TsFile——一种为 IoT 时序数据设计的高性能列式文件格式。这一方案能够实现高效的本地存储、压缩，以及与云端 Apache IoTDB 的无缝数据同步。与会者将了解该工具的架构，观看从模拟机器人进行边缘数据采集的现场演示，并理解如何将这一模式扩展到机器人与工业 IoT 领域，构建完整的边云协同数据管道。 + +### 讲师: + + +
+ +Ansi Zhang: ROS + Apache TsFile | Edge Data Management + +Ansi Zhang 是贵州大学副教授，同时也是清华大学的访问学者。他的研究方向聚焦于无人机大数据和工业大数据，这些领域对高效的边缘数据管理有着极高要求。他主持了 7 项国家级和省级科研项目，发表 20 余篇 SCI/EI 论文，拥有 19 项专利和 10 项软件著作权。作为 ROS-to-TsFile 工具的创建者，他致力于打通机器人数据采集与 Apache 大数据生态之间的鸿沟，为边缘侧的自主系统实现轻量化、高压缩比的存储。在 Community Over Code 大会上，他期待分享 Apache TsFile 如何变革机器人和工业 IoT 领域的数据管道。 \ No newline at end of file diff --git a/content/sessions/keynote-1214444.md b/content/sessions/keynote-1214444.md new file mode 100644 index 000000000..a9dfb1d0b --- /dev/null +++ b/content/sessions/keynote-1214444.md @@ -0,0 +1,68 @@ +--- +title: "Empowering Agentic AI: How Apache Projects Can Shape the Next Open Data and AI Stack" +date: "2026-08-07T11:30:00" +track: "keynote" +presenters: "Junping Du, William Guo, Jerry Shao, Tom Tan, Attila Turoczy, Mark Hoerth" +stype: "English Session" +room: "MainRoom - YiHe Hall" +--- + +As AI systems evolve from passive assistants into more autonomous, goal-driven agents, the need for open, reliable, and interoperable infrastructure becomes increasingly important. This panel explores how Apache projects can help empower the next wave of agentic AI by providing the core building blocks for data access, processing, governance, streaming, orchestration, and execution. + +From Apache Kafka for real-time event streams, Apache Spark and Apache Flink for large-scale data processing, Apache Iceberg for modern lakehouse storage, to metadata and governance projects such as Apache Gravitino, the Apache ecosystem already provides much of the foundation needed for intelligent agents to interact with enterprise data safely and effectively. The discussion will examine where open-source data infrastructure is already enabling agentic workflows, what technical gaps still remain, and how Apache communities can collaborate to define the next generation of AI-native data systems. + +This panel is both a reflection on Apache’s long-standing role in shaping the modern data stack and a forward-looking conversation about how open communities can help build trusted, governed, and extensible foundations for agentic AI. + +### Speakers: + + +
+ +Junping Du: Found & CEO + +Founder and CEO of Datastrato, Ex-Chairperson of LF AI & DATA, ASF Member, Committer and PMC for Apache Hadoop, Co-founder of Apache Ozone, YuniKorn, etc. + + +
+ +William Guo: Apache Software Foundation Member + +* Apache Software Foundation Member +* Apache IPMC Member +* PMC of Apache DolphinScheduler +* Mentor of Apache SeaTunnel(incubating) +* Founder of ClickHouse China Community +* Track Chair of Workflow/Data Governance of Apache Con Asia 2021/2022 + +William used to be the CTO of Analysys and the Senior Big Data Director of Lenovo, general manager of bid data in Wanda. He worked as Big Data Director/manager at CICC, IBM, and Teradata. He has more than 20 years of experience in big data technology and data management. + + +
+ +Jerry Shao: Datastrato, CTO + +Jerry Shao is the co-founder and CTO of Datastrato, focused on open source Big Data are for more than 10 years. He is an Apache member, committer and PMC member of Apache Spark and Apache Inlong, the original creator of Apache Gravitino. + + +
+ +Tom Tan: Datastrato advisor + +Advisor Datastrato & Kwaai AI lab +Head of AI, data and infra, SmartNews +VP of engineering, cloudwalk +Director, Apple data and ML platform + + +
+ +Mark Hoerth: Product Lead, Datastrato + +Mark Hoerth is Product Lead and Solutions Architect at Datastrato, where he works with enterprise data teams to turn federated governance requirements into shipped product, most recently Iceberg REST Catalog federation and multi-cloud credential vending in Gravitino 1.3. He previously spent seven years at Dremio, since acquired by SAP, ending as Principal Product Manager for core technology including Apache Iceberg and Dremio's Polaris-based open catalog. He holds two degrees from Stanford University and is based in the San Francisco Bay Area. + + +
+ +Attila Turóczy: Senior Director of Engineering at Cloudera + +Apache Hive, Impala and Big Data enthusiasm at Cloudera \ No newline at end of file diff --git a/content/sessions/keynote-1214444.zh.md b/content/sessions/keynote-1214444.zh.md new file mode 100644 index 000000000..ffe081f25 --- /dev/null +++ b/content/sessions/keynote-1214444.zh.md @@ -0,0 +1,68 @@ +--- +title: "赋能智能体 AI：Apache 项目如何塑造下一代开放数据与 AI 技术栈" +date: "2026-08-07T11:30:00" +track: "keynote" +presenters: "Junping Du, William Guo, Jerry Shao, Tom Tan, Attila Turoczy, Mark Hoerth" +stype: "英文演讲" +room: "主会场 - 颐和厅" +--- + +随着 AI 系统从被动的助手演变为更加自主、目标驱动的智能体，对开放、可靠且可互操作的基础设施的需求变得愈发重要。本场座谈将探讨 Apache 项目如何通过提供数据访问、处理、治理、流式处理、编排与执行的核心构建模块，助力赋能下一波智能体 AI。 + +从用于实时事件流的 Apache Kafka、用于大规模数据处理的 Apache Spark 与 Apache Flink、用于现代湖仓存储的 Apache Iceberg，再到 Apache Gravitino 等元数据与治理项目，Apache 生态已经提供了智能体安全、高效地与企业数据交互所需的大部分基础。讨论将审视开源数据基础设施已在哪些地方支撑起智能体工作流、目前仍存在哪些技术缺口，以及 Apache 社区如何协作来定义下一代 AI 原生的数据系统。 + +本场座谈既是对 Apache 在塑造现代数据技术栈中长期作用的回顾，也是一场面向未来的对话——探讨开放社区如何助力为智能体 AI 构建可信、受治理且可扩展的基石。 + +### 讲师: + + +
+ +Junping Du：创始人兼 CEO + +Datastrato 创始人兼 CEO，LF AI & DATA 前主席，ASF 成员，Apache Hadoop Committer 与 PMC 成员，Apache Ozone、YuniKorn 等项目的联合创始人。 + + +
+ +William Guo：Apache 软件基金会成员 + +* Apache 软件基金会成员 +* Apache IPMC 成员 +* Apache DolphinScheduler PMC 成员 +* Apache SeaTunnel（孵化中）导师 +* ClickHouse 中国社区创始人 +* ApacheCon Asia 2021/2022 工作流/数据治理分会主席 + +William 曾任易观（Analysys）CTO、联想大数据高级总监、万达大数据总经理。他曾在中金公司（CICC）、IBM 和 Teradata 担任大数据总监/经理。他在大数据技术与数据管理领域拥有超过 20 年的经验。 + + +
+ +Jerry Shao：Datastrato CTO + +Jerry Shao 是 Datastrato 的联合创始人兼 CTO，专注于开源大数据领域已超过 10 年。他是 Apache 成员、Apache Spark 与 Apache Inlong 的 Committer 和 PMC 成员，也是 Apache Gravitino 的原创始人。 + + +
+ +Tom Tan：Datastrato 顾问 + +Datastrato 与 Kwaai AI 实验室顾问 +SmartNews AI、数据与基础设施负责人 +云从科技（CloudWalk）工程副总裁 +苹果数据与机器学习平台总监 + + +
+ +Mark Hoerth：产品负责人，Datastrato + +Mark Hoerth 是 Datastrato 的产品负责人与解决方案架构师。他与企业数据团队合作，将联邦治理需求转化为实际产品，近期成果包括 Gravitino 1.3 中的 Iceberg REST Catalog 联邦与多云凭证分发（credential vending）。此前他在 Dremio 工作了七年（该公司后被 SAP 收购），最终担任核心技术首席产品经理，负责范围包括 Apache Iceberg 与 Dremio 基于 Polaris 的开放目录。他拥有斯坦福大学的两个学位，现居旧金山湾区。 + + +
+ +Attila Turóczy：Cloudera 工程高级总监 + +在 Cloudera 致力于 Apache Hive、Impala 及大数据领域的推广。 \ No newline at end of file diff --git a/content/sessions/keynote-1228764.md b/content/sessions/keynote-1228764.md new file mode 100644 index 000000000..5ae937c5e --- /dev/null +++ b/content/sessions/keynote-1228764.md @@ -0,0 +1,32 @@ +--- +title: "Harness Engineering for DataOps" +date: "2026-08-08T10:30:00" +track: "keynote" +presenters: "William Guo" +stype: "Chinese Session" +room: "MainRoom - YiHe Hall" +--- + +Over the past decade, the Modern Data Stack has built a rich ecosystem of tools for data engineers, covering data ingestion, lakehouse architecture, orchestration, governance, BI, and AI. However, in the Agentic era, data platforms need to evolve further: from “humans operating tools” to “humans defining goals, and Agents orchestrating execution.” + +This talk will explore a new platform paradigm: Harness Engineering for Data. It will discuss how data ingestion, ETL/ELT, SQL development, orchestration, data quality, metadata, lineage, monitoring, and enterprise ontology can be unified and packaged as engineering capabilities that Agents can understand, invoke, and govern. + +By injecting business ontology, technical metadata, and execution context into Agent workflows, enterprises can enable Agents to better understand data objects and business relationships, execute Skills more efficiently, reduce context guessing and token consumption, and improve the accuracy of ETL pipeline generation, SQL output, data modeling, and visualization analysis. + +Ultimately, data engineering platforms will evolve from traditional collections of tools into a Data Engineering Harness designed for human-Agent collaboration, helping enterprises achieve more efficient, trustworthy, and governable data engineering delivery in the Agentic era. + +### Speakers: + + +
+ +William Guo: Apache Software Foundation Member + +* Apache Software Foundation Member +* Apache IPMC Member +* PMC of Apache DolphinScheduler +* Mentor of Apache SeaTunnel(incubating) +* Founder of ClickHouse China Community +* Track Chair of Workflow/Data Governance of Apache Con Asia 2021/2022 + +William used to be the CTO of Analysys and the Senior Big Data Director of Lenovo, general manager of bid data in Wanda. He worked as Big Data Director/manager at CICC, IBM, and Teradata. He has more than 20 years of experience in big data technology and data management. \ No newline at end of file diff --git a/content/sessions/keynote-1228764.zh.md b/content/sessions/keynote-1228764.zh.md new file mode 100644 index 000000000..199e346b3 --- /dev/null +++ b/content/sessions/keynote-1228764.zh.md @@ -0,0 +1,32 @@ +--- +title: "面向 DataOps 的 Harness 工程化" +date: "2026-08-08T10:30:00" +track: "keynote" +presenters: "William Guo" +stype: "中文演讲" +room: "主会场 - 颐和厅" +--- + +在过去十年里，现代数据栈（Modern Data Stack）为数据工程师构建了一个丰富的工具生态，覆盖数据接入、湖仓架构、编排、治理、BI 和 AI。然而在 Agentic（智能体）时代，数据平台需要进一步演进：从"人类操作工具"，走向"人类定义目标、由 Agent 编排执行"。 + +本次演讲将探讨一种新的平台范式：面向数据的 Harness 工程（Harness Engineering for Data）。它将讨论如何把数据接入、ETL/ELT、SQL 开发、编排、数据质量、元数据、血缘、监控，以及企业本体（ontology）统一起来，封装成 Agent 能够理解、调用并治理的工程能力。 + +通过将业务本体、技术元数据和执行上下文注入到 Agent 的工作流中，企业可以让 Agent 更好地理解数据对象和业务关系，更高效地执行 Skills，减少对上下文的猜测以及 token 消耗，并提升 ETL 管道生成、SQL 产出、数据建模和可视化分析的准确性。 + +最终，数据工程平台将从传统的工具集合，演进为一种面向人-Agent 协作而设计的 Data Engineering Harness，帮助企业在 Agentic 时代实现更高效、更可信、更可控的数据工程交付。 + +### 讲师: + + +
+ +William Guo：Apache Software Foundation Member + +* Apache Software Foundation Member +* Apache IPMC Member +* Apache DolphinScheduler PMC +* Apache SeaTunnel（孵化中）Mentor +* ClickHouse 中国社区创始人 +* ApacheCon Asia 2021/2022 工作流/数据治理 Track Chair + +William 曾担任易观（Analysys）的 CTO、联想的大数据高级总监，以及万达的大数据总经理。他曾在 CIBM、IBM 和 Teradata 担任大数据总监/经理。他在大数据技术与数据管理领域拥有 20 年以上的经验。 \ No newline at end of file diff --git a/content/sessions/keynote-1275170.md b/content/sessions/keynote-1275170.md new file mode 100644 index 000000000..a836505bf --- /dev/null +++ b/content/sessions/keynote-1275170.md @@ -0,0 +1,19 @@ +--- +title: "The State of Open Source AI in China and Global Implications（中国开源人工智能发展现状与全球影响力）" +date: "2026-08-09T09:00:00" +track: "keynote" +presenters: "Wei Wang" +stype: "Chinese Session" +room: "MainRoom - YiHe Hall" +--- + +Artificial intelligence is reshaping the global technological landscape at an unprecedented pace, with open source serving as the core engine driving AI innovation and democratization. Yet beneath this prosperity lie profound challenges: while generative AI significantly boosts individual developer productivity, it simultaneously erodes the collaborative culture essential to open source communities; meanwhile, the explosive growth of open-source large models has triggered new issues like license conflicts and lagging governance frameworks. Drawing on key insights from the 2025 China Open Source Annual Report, this talk systematically analyzes the current state, key drivers, and structural bottlenecks of China’s open-source AI ecosystem. It addresses a fundamental question for the AI-native era: how to redesign collaboration mechanisms to ensure open source remains not just "code publication," but genuine "collaboration in action." The presentation will focus on building future-oriented governance frameworks for open-source AI, advocating for the inclusion of large models and other novel artifacts into the global public goods system, and exploring how standards, contribution evaluation, and community practices can bridge national strategies, industrial needs, and global open-source governance—to provide actionable pathways for deepening the "Community Over Code" ethos in the age of AI. + +### Speakers: + + +
+ +Wei Wang: Professor of East China Normal University + +Wang Wei, currently a professor at the School of Data Science and Engineering, East China Normal University, X-lab open laboratory founder. Outstanding member of CCF, member of CCF Open Source Development Committee, person in charge of open source governance series standards of China Electronics Standardization Institute, member of Mulan Open Source Community Technical Committee, director of Open Source Club. His research interests is open source development ecology. \ No newline at end of file diff --git a/content/sessions/keynote-1275170.zh.md b/content/sessions/keynote-1275170.zh.md new file mode 100644 index 000000000..00f03fd78 --- /dev/null +++ b/content/sessions/keynote-1275170.zh.md @@ -0,0 +1,19 @@ +--- +title: "中国开源人工智能发展现状与全球影响力" +date: "2026-08-09T09:00:00" +track: "keynote" +presenters: "Wei Wang" +stype: "中文演讲" +room: "主会场 - 颐和厅" +--- + +人工智能正以前所未有的速度重塑全球技术格局，而开源正是驱动 AI 创新与民主化的核心引擎。然而，繁荣之下潜藏着深层挑战：生成式 AI 在显著提升个人开发者生产力的同时，也在侵蚀开源社区赖以维系的协作文化；与此同时，开源大模型的爆发式增长引发了许可证冲突、治理框架滞后等新问题。本次演讲将基于《2025 中国开源年度报告》的核心洞察，系统分析中国开源 AI 生态的现状、关键驱动力与结构性瓶颈。它将回应 AI 原生时代的一个根本问题：如何重新设计协作机制，以确保开源不只是"代码发布"，而是真正的"协作在行动"。演讲将聚焦于构建面向未来的开源 AI 治理框架，倡导将大模型等新型产物纳入全球公共品体系，并探讨标准、贡献评估与社区实践如何连接国家战略、产业需求与全球开源治理--为在 AI 时代深化"Community Over Code（社区重于代码）"精神提供可落地的路径。 + +### 讲师: + + +
+ +Wei Wang：华东师范大学教授 + +王伟，现任华东师范大学数据科学与工程学院教授，X-lab 开放实验室创始人。CCF 杰出会员，CCF 开源发展委员会成员，中国电子技术标准化研究院开源治理系列标准负责人，木兰开源社区技术委员会委员，开源社理事。其研究方向为开源发展生态。 \ No newline at end of file diff --git a/content/sessions/keynote-1277291.md b/content/sessions/keynote-1277291.md new file mode 100644 index 000000000..ac1ce5830 --- /dev/null +++ b/content/sessions/keynote-1277291.md @@ -0,0 +1,21 @@ +--- +title: "The Infrastructure of Trust: Preparing the Apache Software Foundation for AI-Assisted Open Source" +date: "2026-08-08T10:00:00" +track: "keynote" +presenters: "Chris Wells" +stype: "English Session" +room: "MainRoom - YiHe Hall" +--- + +For over two decades, the Apache Software Foundation has maintained the technical backbone that powers hundreds of open source projects used by billions of people worldwide. The ASF Infrastructure team operates the systems that make this possible: the CI pipelines, identity systems, security tooling, and governance frameworks that keep contribution flowing at scale. AI coding assistants are now rapidly lowering the barrier to open source participation, bringing a significant increase in contributors, pull requests, and code that is authored or assisted by automated tools. + +This talk explores that change and the questions it raises for the infrastructure that supports open source contribution. Open source has always depended on trust in contributor identity, code review, and the long-term maintainability of what gets merged. The discussion will cover the role of automated quality gates, CI hardening, and code scanning at scale; the importance of maintaining clear human ownership and accountability for every contribution, even when that contribution is AI-assisted; and the broader responsibility that comes with supporting infrastructure so many projects depend on. The goal is not to resist AI-assisted contribution, but to consider how the systems that validate, verify, and govern it can be ready for what is coming. + +### Speakers: + + +
+ +Chris Wells: Infra Nerd + +Chris Wells is an Infrastructure Engineer and Member of the Apache Software Foundation. He has worked in technology for nearly thirty years, beginning as a teenager who kept the computers running at his local public library. Since then he has been a library IT manager, a database administrator, a consultant, a publisher, and a nonprofit co-founder. Today he helps operate the infrastructure behind the Foundation's projects and contributors. His work spans observability, networking, automation, and AI-assisted development. Chris believes technology should make knowledge easier to reach for everyone, and he builds with that goal in mind. \ No newline at end of file diff --git a/content/sessions/keynote-1277291.zh.md b/content/sessions/keynote-1277291.zh.md new file mode 100644 index 000000000..6ef64e9a0 --- /dev/null +++ b/content/sessions/keynote-1277291.zh.md @@ -0,0 +1,21 @@ +--- +title: "信任的基础设施：让 Apache 软件基金会为 AI 辅助开源做好准备" +date: "2026-08-08T10:00:00" +track: "keynote" +presenters: "Chris Wells" +stype: "英文演讲" +room: "主会场 - 颐和厅" +--- + +二十多年来，Apache 软件基金会一直维系着支撑数百个开源项目运转的技术骨干，这些项目为全球数十亿人所使用。ASF 基础设施团队负责运维让这一切成为可能的系统：CI 流水线、身份系统、安全工具以及治理框架，确保大规模的贡献得以持续流动。如今，AI 编程助手正迅速降低参与开源的门槛，带来了贡献者、拉取请求以及由自动化工具编写或辅助编写的代码的显著增长。 + +本次演讲将探讨这一变化，以及它为支撑开源贡献的基础设施所引发的诸多问题。开源一直以来都依赖于对贡献者身份、代码审查以及所合并内容长期可维护性的信任。讨论将涵盖自动化质量门禁、CI 加固与大规模代码扫描所扮演的角色；即便贡献是由 AI 辅助完成的，为每一项贡献保持清晰的人类归属与问责的重要性；以及支撑如此众多项目所依赖的基础设施所带来的更广泛责任。目标并非抗拒 AI 辅助的贡献，而是思考那些用于验证、核实与治理它的系统，如何才能为即将到来的浪潮做好准备。 + +### 讲师: + + +
+ +Chris Wells：基础设施极客 + +Chris Wells 是一名基础设施工程师，也是 Apache 软件基金会的成员。他在技术领域工作了近三十年，最初还是个少年时，他就在当地的公共图书馆维护那些计算机的运转。此后，他先后担任过图书馆 IT 主管、数据库管理员、顾问、出版人以及一家非营利机构的联合创始人。如今，他协助运维支撑基金会各项目与贡献者的基础设施。他的工作横跨可观测性、网络、自动化以及 AI 辅助开发。Chris 相信技术应当让知识对每一个人都更易触达，他也始终以此为目标来进行构建。 \ No newline at end of file diff --git a/content/sessions/keynote-1285193.md b/content/sessions/keynote-1285193.md new file mode 100644 index 000000000..74e09c7dc --- /dev/null +++ b/content/sessions/keynote-1285193.md @@ -0,0 +1,19 @@ +--- +title: "How to Use Scaffolding to Make Your Coding Agents Less Dumb" +date: "2026-08-09T09:30:00" +track: "keynote" +presenters: "Quincy Larson" +stype: "English Session" +room: "MainRoom - YiHe Hall" +--- + +Learn how scaffolding can speed up development and reduce your token usage. I'll walk you through the key decisions you should make before sending agents into your codebase. Then I'll demo freeCodeCamp’s new open source create-constellation-app tool. + +### Speakers: + + +
+ +Quincy Larson: Founder of freeCodeCamp.org + +Quincy Larson is a teacher and founder of freeCodeCamp.org, an open source project that has helped helped millions of people learn programming. \ No newline at end of file diff --git a/content/sessions/keynote-1285193.zh.md b/content/sessions/keynote-1285193.zh.md new file mode 100644 index 000000000..e2140111d --- /dev/null +++ b/content/sessions/keynote-1285193.zh.md @@ -0,0 +1,19 @@ +--- +title: "如何用脚手架让你的编程 Agent 不再那么笨" +date: "2026-08-09T09:30:00" +track: "keynote" +presenters: "Quincy Larson" +stype: "英文演讲" +room: "主会场 - 颐和厅" +--- + +了解脚手架如何加速开发并降低你的 token 用量。我将带你梳理在把 Agent 放进你的代码库之前应该做出的几个关键决策。随后，我会演示 freeCodeCamp 全新的开源工具 create-constellation-app。 + +### 讲师: + + +
+ +Quincy Larson：freeCodeCamp.org 创始人 + +Quincy Larson 是一名教师，也是 freeCodeCamp.org 的创始人--这是一个帮助了数百万人学习编程的开源项目。 \ No newline at end of file diff --git a/content/sessions/keynote-1288711.md b/content/sessions/keynote-1288711.md new file mode 100644 index 000000000..693fa4ea0 --- /dev/null +++ b/content/sessions/keynote-1288711.md @@ -0,0 +1,20 @@ +--- +title: "Community Over Code in the Era of Agentic Coding: Fostering Trust and Consensus" +date: "2026-08-08T09:30:00" +track: "keynote" +presenters: "Zili Chen" +stype: "Chinese Session" +room: "MainRoom - YiHe Hall" +--- + +Agentic coding lowers contribution barriers and encourages lightweight forks. This session explores why ASF communities, consensus, and trust remain the essential bedrock for verifiable software when AI writes the code. + + +### Speakers: + + +
+ +Zili Chen: Co-Founder, ScopeDB + +Zili Chen (@tison) co-founded ScopeDB, a database built for semi-structured data with cloud services. He is a seasoned open-source advocate and ASF member with a wealth of experience. He has maintained several significant projects, including Apache Flink, Apache ZooKeeper, and Apache Pulsar. \ No newline at end of file diff --git a/content/sessions/keynote-1288711.zh.md b/content/sessions/keynote-1288711.zh.md new file mode 100644 index 000000000..12e3fbfe8 --- /dev/null +++ b/content/sessions/keynote-1288711.zh.md @@ -0,0 +1,20 @@ +--- +title: "Agentic Coding 时代的 Community Over Code：培育信任与共识" +date: "2026-08-08T09:30:00" +track: "keynote" +presenters: "Zili Chen" +stype: "中文演讲" +room: "主会场 - 颐和厅" +--- + +Agentic coding 降低了贡献的门槛，也催生了更多轻量级的 fork。本次分享将探讨：当 AI 来编写代码时，为何 ASF 社区、共识与信任依然是可验证软件不可或缺的基石。 + + +### 讲师: + + +
+ +Zili Chen：ScopeDB 联合创始人 + +陈梓立（@tison）联合创立了 ScopeDB，一款面向半结构化数据、内置云服务的数据库。他是一位资深的开源倡导者，也是 ASF 成员，经验丰富。他曾维护过多个重要项目，包括 Apache Flink、Apache ZooKeeper 和 Apache Pulsar。 \ No newline at end of file diff --git a/content/sessions/keynote-1290495.md b/content/sessions/keynote-1290495.md new file mode 100644 index 000000000..c698f2c0c --- /dev/null +++ b/content/sessions/keynote-1290495.md @@ -0,0 +1,25 @@ +--- +title: "Many Eyes No Longer: Open Source Security in the Age of AI" +date: "2026-08-07T10:00:00" +track: "keynote" +presenters: "Jarek Potiuk" +stype: "English Session" +room: "MainRoom - YiHe Hall" +--- + +For twenty-five years, open source security rested on two assumptions: that enough human eyes would eventually catch any bug, and that the people showing up to help could, over time, be trusted. AI dissolves both. A single actor can now generate an endless stream of plausible patches, file superhuman-polished vulnerability reports, and play the patient, helpful long-term contributor — the exact social-engineering pattern behind the xz backdoor — at a scale no volunteer maintainer can match. The attackers already have their force multiplier. Most maintainers do not (yet). + +This talk examines what open source security actually looks like when both sides of a commit might be a machine: AI "slop" reports drowning security inboxes, and self-propagating supply-chain worms like Shai-Hulud moving faster than any human can triage. Drawing on a decade of building tooling for maintainers — from Breeze to the agentic workflows behind Magpie — it argues the answer isn't asking exhausted volunteers to work harder. It's handing them the same asymmetric leverage attackers already wield: defensive, agentic infrastructure that protects the one thing AI can't replace — human trust — by guarding the scarce human attention that grants it. + +### Speakers: + + +
+ +Jarek Potiuk: Independent Open-Source Contributor and Advisor + +Independent Open-Source Contributor and Advisor, Committer and PMC member of Apache Airflow, Member of the Apache Software Foundation + +Jarek is an Engineer with a broad experience in many subjects - Open-Source, Cloud, Mobile, Robotics, AI, Backend, Developer Experience, but he also had a lot of non-engineering experience - running a company, being CTO, organizing big, international community events, technical sales support, pr and marketing advisory but also looking at legal aspect of licensing and building open-source communities are all under his belt. + +With the experience in very small and very big companies and everything in-between, Jarek found his place in Open-Source world, where his internal individual-contributor drive can be used to the uttermost of the potential. \ No newline at end of file diff --git a/content/sessions/keynote-1290495.zh.md b/content/sessions/keynote-1290495.zh.md new file mode 100644 index 000000000..932a81095 --- /dev/null +++ b/content/sessions/keynote-1290495.zh.md @@ -0,0 +1,25 @@ +--- +title: "再无众目：AI 时代的开源安全" +date: "2026-08-07T10:00:00" +track: "keynote" +presenters: "Jarek Potiuk" +stype: "英文演讲" +room: "主会场 - 颐和厅" +--- + +二十五年来，开源安全建立在两个假设之上：足够多的人眼终将发现任何 bug，以及那些前来帮忙的人，假以时日，是可以被信任的。AI 把这两点都瓦解了。如今，单一行为者就能生成源源不断、看似合理的补丁，提交打磨得超乎人类水平的漏洞报告，并扮演一个耐心、乐于助人的长期贡献者--这正是 xz 后门背后的那套社会工程学手法--而且规模之大，没有任何志愿维护者能够匹敌。攻击者已经握有他们的力量倍增器。而大多数维护者（还）没有。 + +本次演讲将审视：当一次提交的双方都可能是一台机器时，开源安全究竟会是什么样子--AI"垃圾"报告淹没安全收件箱，像 Shai-Hulud 这样自我传播的供应链蠕虫蔓延速度远超任何人类分诊所能及。基于十年间为维护者构建工具的经验--从 Breeze 到 Magpie 背后的 agentic 工作流--它论证：答案并非要求疲惫的志愿者更拼命地工作，而是赋予他们攻击者业已握有的同一种非对称杠杆--防御性的、agentic 的基础设施，它通过守护那授予信任的稀缺人类注意力，来保护 AI 无法取代的那样东西：人类的信任。 + +### 讲师: + + +
+ +Jarek Potiuk：独立开源贡献者与顾问 + +独立开源贡献者与顾问，Apache Airflow Committer 与 PMC 成员，Apache 软件基金会成员 + +Jarek 是一位工程师，在众多领域拥有丰富经验--开源、云计算、移动端、机器人、人工智能、后端、开发者体验；同时他也有大量非工程领域的经验--经营公司、担任 CTO、组织大型国际社区活动、技术销售支持、公关与营销咨询，以及研究许可证的法律层面和构建开源社区，这些都尽在他的掌握之中。 + +凭借在从极小到极大以及介于两者之间的各类公司中积累的经验，Jarek 在开源世界中找到了自己的位置，在这里，他作为个人贡献者的内在驱动力能够被发挥到极致。 \ No newline at end of file diff --git a/content/sessions/keynote-1296710.md b/content/sessions/keynote-1296710.md new file mode 100644 index 000000000..ddfc8a8ce --- /dev/null +++ b/content/sessions/keynote-1296710.md @@ -0,0 +1,29 @@ +--- +title: "Open Source in the Age of AI Agents: Ecosystem Shifts and Opportunities for China " +date: "2026-08-07T09:30:00" +track: "keynote" +presenters: "Tao Jiang" +stype: "Chinese Session" +room: "MainRoom - YiHe Hall" +--- + +This is the best of times for open source: a weekend project can top GitHub in just 60 days, open-weight models are rapidly closing the gap with proprietary flagships, and an open-source inference engine has raised the largest seed round in history. + +It is also the worst of times for open source: maintainers are drowning in AI-generated junk pull requests, bounty programs are shutting down, and the most valuable engineering data is bypassing GitHub altogether. + +This talk advances a central argument: agents are not here simply to write more code for open source. They are changing the foundation on which open source has operated for the past thirty years—code is becoming cheaper, while trust is becoming more expensive. + +Open source must now rewrite its answers to four fundamental questions: Why is code valuable? Why do developers contribute? How do companies make money? And who defines the rules? + +A new layer will be built on top of today’s collaboration platforms, with identity, memory, reputation, and settlement forming a new institutional stack. + +For the first time in this transformation, China is not merely a follower. The race in open-weight models, the divergence of two technology ecosystems, the open space for agent collaboration platforms, and the window for neutral governance represent four major opportunities for China’s open-source community. + +### Speakers: + + +
+ +Tao Jiang: Founder and Chairman of CSDN + +Founder and Chairman of CSDN. Currently serves as Vice Chairman of the China Open Source Alliance, Director of the Intelligent Application Branch of the China Software Industry Association, Board Member of the OpenAtom Foundation, and Chairman of the Singularity Intelligence Research Institute. \ No newline at end of file diff --git a/content/sessions/keynote-1296710.zh.md b/content/sessions/keynote-1296710.zh.md new file mode 100644 index 000000000..7a97c5586 --- /dev/null +++ b/content/sessions/keynote-1296710.zh.md @@ -0,0 +1,19 @@ +--- +title: "Agent 时代的开源生态与中国机会" +date: "2026-08-07T09:30:00" +track: "keynote" +presenters: "Tao Jiang" +stype: "中文演讲" +room: "主会场 - 颐和厅" +--- + +这是开源最好的时刻:周末项目 60 天登顶 GitHub,开放权重模型直追闭源旗舰,开源推理引擎拿到史上最大种子轮。这也是开源最坏的时刻:AI 垃圾 PR 淹没维护者,赏金关停,最值钱的工程数据绕开了 GitHub。本演讲提出一个判断:Agent 不是来替开源写更多代码的,它在换开源三十年赖以成立的地基——代码变便宜,信任变昂贵。开源的四个答案(代码为何值钱、开发者为何贡献、公司怎么赚钱、秩序谁来定义)都要重写;协作平台将被"再盖一层",身份、记忆、信誉、结算构成新的制度栈。而这一轮,中国第一次不是跟随者:开放权重的追击、双生态的分化、Agent 协作平台的空位与中立治理的窗口,是摆在中国开源社区面前的四个机会。 + +### 讲师: + + +
+ +Tao Jiang: CSDN创始人、董事长 + +CSDN创始人、董事长，现任中国开源联盟副主席、中国软件协会智能应用分会主任委员，开放原子开源基金会理事、奇点智能研究院理事长 。 \ No newline at end of file diff --git a/content/sessions/keynote-1296719.md b/content/sessions/keynote-1296719.md new file mode 100644 index 000000000..0da29d68d --- /dev/null +++ b/content/sessions/keynote-1296719.md @@ -0,0 +1,69 @@ +--- +title: "How Should Open Source Communities Govern AI-Generated Code" +date: "2026-08-08T11:00:00" +track: "keynote" +presenters: "Willem Jiang, Jarek Potiuk, Craig Russell, Justin Mclean, Zili Chen" +stype: "English Session" +room: "MainRoom - YiHe Hall" +--- + +Open-source communities have traditionally worked to attract more contributions. Agentic coding reverses that assumption: code, documentation, issues, pull requests, and even security reports can now be produced at a scale that volunteer reviewers were never designed to absorb. + +When producing something that looks like a contribution becomes almost free, how should a community decide what deserves attention, who is accountable, and who earns trust? Should AI use be disclosed? Should the ASF define a common baseline, or should every project decide independently? Can an agent ever become a contributor in the community sense? + +Moderated by Willem Jiang, this panel brings together Jarek Potiuk, Craig Russell, Zili Chen, and Justin McLean to debate AI-generated contributions from the perspectives of security, foundation governance, community consensus, and mentoring. The session will be conducted in English. + +### Speakers: + + +
+ +Willem Jiang: Independent Open Source and Agentic AI Consultant + +Willem Jiang (Jiang Ning) is an independent open source and Agentic AI consultant and a maintainer of DeerFlow. He previously served as an Open Source Evangelist at ByteDance’s OSPO and as a member of the Board of Directors of the Apache Software Foundation. +Willem has contributed to numerous Apache projects, including Apache Camel, CXF, ServiceMix, and ServiceComb. Before joining ByteDance, he spent six years at Huawei as an open source technical expert. Earlier in his career, he was a Principal Engineer at Red Hat, where he worked on Fuse ESB, and held engineering roles at FuseSource, IONA, and TravelSky. + + +
+ +Jarek Potiuk: Independent Open-Source Contributor and Advisor + +Independent Open-Source Contributor and Advisor, Committer and PMC member of Apache Airflow, Member of the Apache Software Foundation + +Jarek is an Engineer with a broad experience in many subjects - Open-Source, Cloud, Mobile, Robotics, AI, Backend, Developer Experience, but he also had a lot of non-engineering experience - running a company, being CTO, organizing big, international community events, technical sales support, pr and marketing advisory but also looking at legal aspect of licensing and building open-source communities are all under his belt. + +With the experience in very small and very big companies and everything in-between, Jarek found his place in Open-Source world, where his internal individual-contributor drive can be used to the uttermost of the potential. + + +
+ +Craig Russell: ASF Member, Incubator PMC member, ASF Board Chair emeritus + +I joined Apache as a committer on the incubating JDO project in 2005 and became an Apache Member in 2007. I was appointed Assistant Secretary in 2009 and Secretary in 2010, where I served until 2019. I was elected to the Board of Directors in 2019 and currently serve as a Director. + +I was originally attracted to Apache by its approach to governance, where the people doing the work decide what direction to take the project. This is exactly the opposite of how most corporations work, where "the smartest people in the +room (managers and executives)" make the decisions. + +Once my own project was accepted into Apache, I looked around to see what else I could contribute to, and joined a number of projects where I could help. I also started looking at the intellectual property model of licensing of contributions and licensing of the end products. That led me to start contributing to the Secretary role, processing the "paperwork" granting IP rights to the Foundation. + +Recently it has come to my attention that people feel that the Foundation has "too many rules" which go against the free expression that people look for. Looking deeper into this issue, I found that there are some common themes that, if understood better, can make the experience easier. + + + +
+ +Justin Mclean: ASF Director, VP ASF Incubator, Datastrato Community Manager + +Justin Mclean is a long-time contributor to the Apache Software Foundation (ASF) and currently serves as VP of the ASF Incubator and a member of the ASF Board of Directors. Over several years, he has contributed to evolving the Incubator’s governance, policies, and training programs, including documenting how projects operate in practice and analysing the Incubator’s history. Through this work, he has supported hundreds of open source projects in building sustainable and independent communities. + +He has extensive experience guiding projects from early incubation through to successful graduation, and has reviewed over a thousand open source releases. His work focuses on open source governance, community health, and the design of systems that enable collaboration at scale. + +Justin regularly speaks at international conferences on topics such as open source governance, licensing, and community building, sharing practical lessons drawn from real-world experience across a wide range of ASF projects and communities. + + + +
+ +Zili Chen: Co-Founder, ScopeDB + +Zili Chen (@tison) co-founded ScopeDB, a database built for semi-structured data with cloud services. He is a seasoned open-source advocate and ASF member with a wealth of experience. He has maintained several significant projects, including Apache Flink, Apache ZooKeeper, and Apache Pulsar. \ No newline at end of file diff --git a/content/sessions/keynote-1296719.zh.md b/content/sessions/keynote-1296719.zh.md new file mode 100644 index 000000000..4a452d37b --- /dev/null +++ b/content/sessions/keynote-1296719.zh.md @@ -0,0 +1,69 @@ +--- +title: "开源社区应如何治理 AI 生成的代码" +date: "2026-08-08T11:00:00" +track: "keynote" +presenters: "Willem Jiang, Jarek Potiuk, Craig Russell, Justin Mclean, Zili Chen" +stype: "英文演讲" +room: "主会场 - 颐和厅" +--- + +开源社区长期面对的问题是“如何吸引更多贡献者”。Agentic Coding 正在反转这个前提：代码、文档、Issue、Pull Request，甚至漏洞报告，都可以由 Agent 以远超人工的速度生成。 + +更多产出不一定意味着更健康的社区。对于依靠志愿者运作的项目，真正稀缺的资源正在变成 Maintainer 的评审注意力、能够承担后果的人类判断，以及通过长期协作建立的信任。 + +一份由 AI 生成、通过全部测试却没有披露来源的 PR，应该接受吗？ASF 是否需要统一的 AI 贡献底线，还是应继续由各项目自主制定规则？当 AI 生成的代码引发安全或许可问题，提交者、评审者、项目流程和工具提供方分别承担什么责任？一个可以持续提交补丁、回应评审并维护 Fork 的 Agent，能否成为社区意义上的 Contributor？ + +本场英文圆桌由 Willem Jiang 主持，Jarek Potiuk、Craig Russell、Zili Chen 与 Justin Mclean 将从开源安全、基金会政策、社区信任和孵化教育四个方向展开直接交锋。我们不会停留在“AI 有利也有弊”，而会尝试给出能够写入项目贡献指南的具体规则。 + +### 讲师: + + +
+ +Willem Jiang: 独立开源与 Agentic AI 顾问 + +Willem Jiang（姜宁）是一名独立的开源与 Agentic AI 顾问，也是 DeerFlow 的维护者。他此前曾任字节跳动 OSPO 的开源布道师，以及 Apache 软件基金会董事会成员。Willem 曾为众多 Apache 项目做出贡献，包括 Apache Camel、CXF、ServiceMix 和 ServiceComb。在加入字节跳动之前，他在华为担任开源技术专家六年。在职业生涯早期，他曾任 Red Hat 首席工程师，参与 Fuse ESB 的开发，并曾在 FuseSource、IONA 和 TravelSky 担任工程职务。 + + +
+ +Jarek Potiuk: 独立开源贡献者与顾问 + +独立开源贡献者与顾问，Apache Airflow Committer 及 PMC 成员，Apache 软件基金会成员。 + +Jarek 是一名工程师，在众多领域拥有丰富经验——开源、云、移动、机器人、AI、后端、开发者体验；同时他也积累了大量非工程类经验——运营公司、担任 CTO、组织大型国际社区活动、技术销售支持、公关与营销咨询，以及从法律视角审视许可授权和构建开源社区。 + +凭借在从极小型到极大型各类公司以及介于两者之间的各种规模企业中的经历，Jarek 在开源世界找到了自己的位置——在那里，他作为个体贡献者的内在驱动力能够得到最大程度的发挥。 + + +
+ +Craig Russell: ASF 成员、孵化器 PMC 成员、ASF 董事会主席（荣休） + +我于 2005 年作为孵化中的 JDO 项目的 Committer 加入 Apache，并于 2007 年成为 Apache 成员。2009 年我被任命为助理秘书，2010 年起担任秘书，直至 2019 年。2019 年我当选董事会成员，目前担任董事一职。 + +最初吸引我加入 Apache 的是它的治理理念——由实际做事的人决定项目的发展方向。这与大多数公司的运作方式恰好相反，在那些公司里，由“房间里最聪明的人（管理者和高管）”来做决策。 + +当我的项目被 Apache 接纳后，我开始四处寻找还可以贡献什么，并加入了若干我能帮上忙的项目。我也开始研究贡献的知识产权许可模型以及最终产品的许可模型。这促使我开始承担秘书职责，处理将知识产权授予基金会的相关“文书工作”。 + +最近我注意到，不少人觉得基金会“规则太多”，这与人们所追求的自由表达相悖。深入审视这个问题后，我发现其中有一些共同的主题，如果能更好地理解它们，就能让整个过程变得更轻松。 + + + +
+ +Justin Mclean: ASF 董事、ASF 孵化器副总裁、Datastrato 社区经理 + +Justin Mclean 是 Apache 软件基金会（ASF）的长期贡献者，目前担任 ASF 孵化器副总裁及 ASF 董事会成员。多年来，他参与了孵化器治理、政策与培训项目的演进，包括记录项目实际运作方式并分析孵化器历史。通过这些工作，他帮助数百个开源项目构建了可持续且独立的社区。 + +他在引导项目从早期孵化到顺利毕业方面经验丰富，并审阅过上千个开源发布版本。他的工作聚焦于开源治理、社区健康以及支持大规模协作的机制设计。 + +Justin 经常在国际会议上发表演讲，主题涵盖开源治理、许可授权与社区建设，分享来自众多 ASF 项目与社区真实经验的实践教训。 + + + +
+ +Zili Chen: ScopeDB 联合创始人 + +Zili Chen（@tison）联合创办了 ScopeDB，这是一款面向半结构化数据、提供云服务的数据库。他是一位经验丰富的开源倡导者及 ASF 成员，曾维护多个重要项目，包括 Apache Flink、Apache ZooKeeper 和 Apache Pulsar。 \ No newline at end of file diff --git a/content/sessions/keynote-1298213.md b/content/sessions/keynote-1298213.md new file mode 100644 index 000000000..92989be97 --- /dev/null +++ b/content/sessions/keynote-1298213.md @@ -0,0 +1,33 @@ +--- +title: "Responsible AI at Apache Software Foundation: Challenges and Opportunities" +date: "2026-08-08T09:00:00" +track: "keynote" +presenters: "Craig Russell" +stype: "English Session" +room: "MainRoom - YiHe Hall" +--- + +From its creation in the last century, the Apache Software Foundation has operated on the principle that communities of people from around the world are the best stewards of projects that benefit the public. The corporation that manages the resources is organized as a bottom-up foundation, with decisions about project management left primarily to the projects themselves. + +This organizational model has worked well, with communities of projects growing from one project to over 300 projects in just over 25 years. With the spectacular rise of AI, the infrastructure of computing has been fundamentally altered. Anyone who ignores the potential effects of AI will be left behind. + +The ASF community has adopted many AI tools recently, and several members are directly involved in creating the environment to help projects integrate these tools into their processes. + +This talk discusses some of the beneficial uses of AI as it relates to project management, as well as some of the disadvantages that should be considered as ASF moves forward in the Age of AI. Two major announcements include VP, Responsible AI, and Project Magpie. + + +### Speakers: + + +
+ +Craig Russell: ASF Member, Incubator PMC member, ASF Board Chair emeritus + +I joined Apache as a committer on the incubating JDO project in 2005 and became an Apache Member in 2007. I was appointed Assistant Secretary in 2009 and Secretary in 2010, where I served until 2019. I was elected to the Board of Directors in 2019 and currently serve as a Director. + +I was originally attracted to Apache by its approach to governance, where the people doing the work decide what direction to take the project. This is exactly the opposite of how most corporations work, where "the smartest people in the +room (managers and executives)" make the decisions. + +Once my own project was accepted into Apache, I looked around to see what else I could contribute to, and joined a number of projects where I could help. I also started looking at the intellectual property model of licensing of contributions and licensing of the end products. That led me to start contributing to the Secretary role, processing the "paperwork" granting IP rights to the Foundation. + +Recently it has come to my attention that people feel that the Foundation has "too many rules" which go against the free expression that people look for. Looking deeper into this issue, I found that there are some common themes that, if understood better, can make the experience easier. \ No newline at end of file diff --git a/content/sessions/keynote-1298213.zh.md b/content/sessions/keynote-1298213.zh.md new file mode 100644 index 000000000..09ead409b --- /dev/null +++ b/content/sessions/keynote-1298213.zh.md @@ -0,0 +1,32 @@ +--- +title: "Apache 软件基金会的负责任 AI：挑战与机遇" +date: "2026-08-08T09:00:00" +track: "keynote" +presenters: "Craig Russell" +stype: "英文演讲" +room: "主会场 - 颐和厅" +--- + +自上世纪成立以来，Apache 软件基金会（ASF）始终秉持这样一个原则：来自世界各地的社区成员，是那些造福公众的项目最理想的管理者。负责管理资源的企业组织以自下而上的基金会形式运作，项目管理的决策权主要下放给项目自身。 + +这种组织模式运作良好，项目社区在短短 25 年多的时间里，从单个项目发展到超过 300 个项目。随着 AI 的迅猛崛起，计算的基础设施已被根本性地改变。任何忽视 AI 潜在影响的人，都将被时代抛在身后。 + +ASF 社区近期采用了众多 AI 工具，其中一些成员直接参与到相关环境的创建中，帮助各个项目将这些工具融入其开发流程。 + +本次演讲将探讨 AI 在项目管理方面的一些有益应用，以及在 ASF 迈入 AI 时代之际应当审慎考量的一些不足之处。本次还将发布两项重要内容：负责任 AI 副总裁（VP, Responsible AI）以及 Magpie 项目。 + + +### 讲师: + + +
+ +Craig Russell：ASF 成员、孵化器 PMC 成员、ASF 董事会名誉主席 + +我于 2005 年作为孵化中的 JDO 项目的 committer 加入 Apache，并在 2007 年成为 Apache 成员。2009 年我被任命为助理秘书（Assistant Secretary），2010 年担任秘书（Secretary），在该职位上服务至 2019 年。2019 年我当选为董事会成员，目前担任董事。 + +我最初被 Apache 吸引，正是因为它的治理理念——由真正做事的人来决定项目的发展方向。这与大多数公司的运作方式截然相反，后者由"房间里最聪明的人（管理者和高管）"来做决策。 + +当自己的项目被 Apache 接纳后，我开始四处寻找还能贡献些什么，并加入了许多我可以提供帮助的项目。我也开始研究贡献代码的知识产权许可模型以及最终产品的许可模式。这让我开始参与到秘书这一角色中，处理那些将知识产权授予基金会的"书面工作"。 + +最近我注意到，有人觉得基金会"规则太多"，这与人们所追求的自由表达相违背。深入审视这一问题后，我发现其中有一些共同的主题，如果能够更好地理解它们，就能让大家在参与时更加顺畅。 \ No newline at end of file diff --git a/content/sessions/keynote-1298341.md b/content/sessions/keynote-1298341.md new file mode 100644 index 000000000..d021b6013 --- /dev/null +++ b/content/sessions/keynote-1298341.md @@ -0,0 +1,29 @@ +--- +title: "The Old Rules of an Open Ecosystem Under the New Trends of Agentic AI" +date: "2026-08-07T11:00:00" +track: "keynote" +presenters: "Xiaoya Xia, Xu Wang" +stype: "Chinese Session" +room: "MainRoom - YiHe Hall" +--- + +Ecosystems are always changing, yet what determines whether a technology can become public infrastructure remains a set of familiar questions. We'll return to the well-known experience of the Apache community to discuss how open processes and incrementally built trust continue to work within the Agentic AI ecosystem. Open models also raise new questions: models extend the scope of what is opened to weights, data, and evaluation — beyond the license, what else needs to be delivered? + +Starting from the updated Agentic AI Landscape, this talk examines the changes across Agent Infra, Model Infra, large models, and reusable knowledge repositories (awesome*). It then places Apache back onto this technical map, exploring how its long-standing accumulation in data and distributed systems occupies a critical position within the new ecosystem. Meanwhile, Ant Group's practice through inclusionAI offers another concrete sample: how models, infrastructure, and industry applications can come together to form an open technology stack that more people can participate in. + +### Speakers: + + +
+ +Xiaoya Xia: Open Source Analyst, Ant Group + +Xiaoya Xia is a member of the Ant Group OSPO, where she focuses on catalyzing open source success through data-driven insights. Before joining Ant Group, Xiaoya was a PhD at East China Normal University (ECNU), where she concentrated on research into open source ecosystem sustainability. + + +
+ +Xu Wang: Ant Group Open Source and Container Infra + +Xu Wang is the vice chair of the Ant Group Open Source Technology Committee and also the head of the Container Infrastructure team. Xu joined Ant Group in 2019, and currently he is responsible for the development of the container-based Cloud-Native infrastructure and the open-source related strategies of Ant Group. +Before joining Ant Group, Xu was the CTO and co-founder of hyper.sh, a startup on secure container runtime technology, and co-launched Kata Containers project in 2017. \ No newline at end of file diff --git a/content/sessions/keynote-1298341.zh.md b/content/sessions/keynote-1298341.zh.md new file mode 100644 index 000000000..e710d86bf --- /dev/null +++ b/content/sessions/keynote-1298341.zh.md @@ -0,0 +1,27 @@ +--- +title: "Agentic AI 新趋势下，开放生态的那些老规矩" +date: "2026-08-07T11:00:00" +track: "keynote" +presenters: "Xiaoya Xia, Xu Wang" +stype: "中文演讲" +room: "主会场 - 颐和厅" +--- + +生态一直在变，而决定一项技术能不能成为公共基础设施的，仍是一些熟悉的问题。我们会回到 Apache 社区熟悉的经验，讨论公开过程和逐步建立的信任，怎样继续作用在 Agentic AI 生态里。开放模型也带来了新的问题：模型把开放对象扩展到了权重、数据与评测，许可证之外还需要交付什么？ +这场分享从更新后的 Agentic AI Landscape 出发，观察 Agent Infra、Model Infra、大模型及可复用知识仓库（awesome*）的变化；再把 Apache 放回这张技术地图，看看在数据与分布式系统上的长期积累处在新生态里怎样的关键位置？而蚂蚁通过 inclusionAI 的实践，提供了另一个具体样本：模型、基础设施和行业应用，可以怎样形成一套更多人能够参与的开放技术栈。 + +### 讲师: + + +
+ +Xiaoya Xia：开源分析师，蚂蚁集团 + +Xiaoya Xia 是蚂蚁集团 OSPO（开源项目办公室）成员，致力于通过数据驱动的洞察推动开源项目的成功。加入蚂蚁集团之前，Xiaoya 在华东师范大学（ECNU）攻读博士，研究方向为开源生态的可持续发展。 + + +
+ +Xu Wang：蚂蚁集团开源与容器基础设施 + +Xu Wang 是蚂蚁集团开源技术委员会副主席，兼任容器基础设施团队负责人。他于 2019 年加入蚂蚁集团，目前负责基于容器的云原生基础设施建设以及蚂蚁集团的开源相关战略。加入蚂蚁集团之前，Xu 曾是 hyper.sh 的联合创始人兼 CTO，这是一家专注于安全容器运行时技术的初创公司，他于 2017 年共同发起了 Kata Containers 项目。 \ No newline at end of file diff --git a/content/sessions/keynote-1298389.md b/content/sessions/keynote-1298389.md new file mode 100644 index 000000000..c09a20155 --- /dev/null +++ b/content/sessions/keynote-1298389.md @@ -0,0 +1,23 @@ +--- +title: "From Models to Agents: How Open Source Powers the Next Wave of AI Innovation" +date: "2026-08-07T10:30:00" +track: "keynote" +presenters: "Betty Zheng" +stype: "Chinese Session" +room: "MainRoom - YiHe Hall" +--- + +The center of gravity in AI has shifted — from standalone models to autonomous agents that reason, act, and collaborate. And at nearly every layer of this new stack, open source is doing the heavy lifting: open table formats and data infrastructure feeding the models, open agent frameworks orchestrating the reasoning, and open protocols letting agents talk to tools and to each other. + +Drawing on real projects across the Apache ecosystem and beyond — from Apache Iceberg for AI-ready data, to open-source agent SDKs like Strands, to emerging standards such as MCP — this keynote traces how a healthy open source community becomes the foundation for production-grade agentic AI. We'll look at what breaks when agents move from demo to production (state, control, observability), how the community is solving it in the open, and where developers can plug in and contribute. + +You'll leave with a clear mental map of the modern AI + open source stack, and concrete ways to build on it — and give back to it. + +### Speakers: + + +
+ +Betty Zheng: Senior Developer + +Yubin Zheng is a seasoned engineer with over 20 years of experience in ICT and digital transformation. She focuses on building real-world AI systems, including agent runtime design, multi-agent coordination, and production deployment. She is an active contributor to developer and open-source communities, and a frequent speaker at cloud-native and AI events. \ No newline at end of file diff --git a/content/sessions/keynote-1298389.zh.md b/content/sessions/keynote-1298389.zh.md new file mode 100644 index 000000000..fb3675328 --- /dev/null +++ b/content/sessions/keynote-1298389.zh.md @@ -0,0 +1,23 @@ +--- +title: "从模型到智能体：开源如何驱动下一波 AI 创新浪潮" +date: "2026-08-07T10:30:00" +track: "keynote" +presenters: "Betty Zheng" +stype: "中文演讲" +room: "主会场 - 颐和厅" +--- + +AI 的重心已经发生了转移——从独立的模型，转向能够推理、行动并协作的自主智能体（autonomous agents）。而在这个新堆栈的几乎每一层，开源都在承担着核心重任：开放的表格式与数据基础设施为模型供给养分，开放的智能体框架编排着推理过程，而开放协议则让智能体得以与工具以及彼此之间进行沟通。 + +本主题演讲将以 Apache 生态乃至更广范围中的真实项目为例——从面向 AI 就绪数据的 Apache Iceberg，到 Strands 这样的开源智能体 SDK，再到 MCP 等新兴标准——梳理一个健康的开源社区如何成为生产级智能体 AI（agentic AI）的根基。我们将探讨当智能体从演示走向生产时会暴露出哪些问题（状态、控制、可观测性），社区又是如何以开源的方式加以解决，以及开发者可以在哪些环节接入并贡献力量。 + +你将带着一幅清晰的现代"AI + 开源"技术栈全景图离开，并获得在此基础上进行构建——并回馈社区——的具体路径。 + +### 讲师: + + +
+ +Betty Zheng：资深开发者 + +Yubin Zheng 是一位资深的工程师，在 ICT 与数字化转型领域拥有超过 20 年的经验。她专注于构建面向真实场景的 AI 系统，工作涵盖智能体运行时设计、多智能体协同以及生产化部署。她活跃于开发者与开源社区，是云原生与 AI 领域各类活动的常邀演讲嘉宾。 \ No newline at end of file diff --git a/content/sessions/keynote-1298466.md b/content/sessions/keynote-1298466.md new file mode 100644 index 000000000..5b1af4191 --- /dev/null +++ b/content/sessions/keynote-1298466.md @@ -0,0 +1,33 @@ +--- +title: "Open Collaboration for Embodied AI - Insights from RoboChallenge Real-Robot Benchmarking" +date: "2026-08-09T10:20:00" +track: "keynote" +presenters: "Emily Chen" +stype: "Chinese Session" +room: "MainRoom - YiHe Hall" +--- + +Embodied AI is rapidly moving from laboratory research toward real-world applications. With the rapid advancement of Vision-Language-Action (VLA) models, robots are evolving from “understanding the world” to “acting in the world.” At the same time, building open, trustworthy, and reproducible evaluation systems, and leveraging open collaboration to accelerate technological progress, have become critical challenges for the development of embodied AI. + +This keynote will present insights from the RoboChallenge 2025 Annual Report, sharing the latest evaluation results of global embodied AI foundation models in real-world robotic environments. It will also highlight RoboChallenge’s latest progress, including real-robot challenges at leading international conferences such as CVPR and ICRA, as well as the launch of EAI Bench at the AI for Good Global Summit in Geneva. + +Based on tens of thousands of real-robot execution trials, the report analyzes the current state and future trends of embodied AI models. Current leading Vision-Language-Action (VLA) models have achieved approximately 62% success rate on the Table30 benchmark, while success rates for complex fine-manipulation tasks remain below 15%, demonstrating that embodied AI is still at a critical stage of advancing from semantic understanding toward robust physical interaction and complex manipulation capabilities. Meanwhile, RoboChallenge has attracted developers from China, the United States, Singapore, Japan, and other regions, with open-source models, open datasets, and open evaluation becoming increasingly important drivers of innovation. + +In the era of embodied AI, open source is evolving from software sharing toward broader open collaboration. The future requires collective efforts to advance open models, datasets, robot interfaces, evaluation benchmarks, and open standards. RoboChallenge aims to connect global developers, robot manufacturers, research institutions, and industry partners to build an open, trustworthy, and reproducible real-robot evaluation ecosystem, enabling more developers to access real-world robotic resources and accelerate embodied AI innovation. + +### Speakers: + + +
+ +Emily Chen: Co-Founder of Kaiyuanshe + +Emily Chen is the Co-founder of Kaiyuanshe and a current Board Director . A long-time advocate, builder, and connector in the global open source community, she has played a key role in shaping open collaboration across Asia and globle. + +Emily served on the GNOME Foundation Board of Directors in 2010 and founded the GNOME.Asia community in 2008, bringing the GNOME.Asia Summit to over ten countries and regions. She was honored with the GNOME Foundation’s highest recognition, the GNOME Pants Award, at GUADEC 2014 in Sweden. + +From 2017 to 2025, she hosted multiple visits of GitHub’s executive team and global community leaders to China, building bridges between international platforms and the Chinese open source ecosystem. She also served as the China representative to the Open Source Initiative (OSI) Alliance and was Deputy Secretary-General of the China OSS Promotion Union in 2016. + +Emily initiated impactful projects such as the Kaiyuanshe China OPenSource Annual Report, COSCon (China Open Source Conference), and the "33 Open Source Pioneers" list. She is a core contributor to Mozilla, a Google Summer of Code mentor, and one of the organizers of the Open Source Congress 2024. Currently, she serves as a TOC Mentor at the OpenAtom Open Source Foundation. + +With nearly two decades of experience in open source community building and global-local collaboration, Emily continues to champion open innovation in the GenAI era. \ No newline at end of file diff --git a/content/sessions/keynote-1298466.zh.md b/content/sessions/keynote-1298466.zh.md new file mode 100644 index 000000000..d4cd21271 --- /dev/null +++ b/content/sessions/keynote-1298466.zh.md @@ -0,0 +1,29 @@ +--- +title: "Open Collaboration for Embodied AI - Insights from RoboChallenge Real-Robot Benchmarking" +date: "2026-08-09T10:20:00" +track: "keynote" +presenters: "Emily Chen" +stype: "中文演讲" +room: "主会场 - 颐和厅" +--- + +具身智能正加速从实验室研究走向真实世界应用。随着 Vision-Language-Action（VLA）模型快速发展，机器人正在从“理解世界”迈向“操作世界”。与此同时，如何建立开放、可信、可复现的评测体系，如何通过开源协作推动技术快速迭代，成为具身智能发展的关键议题。 本次演讲将基于 RoboChallenge 平台，分享全球具身智能基础模型在真实机器人环境中的最新评测成果，并介绍 RoboChallenge 在 CVPR、ICRA 国际真机挑战赛，以及 AI for Good 全球峰会发布 EAI Bench 等方面的最新进展。 基于数万次真实机器人执行数据，将分析具身智能模型的发展趋势。当前领先 VLA 模型在 Table30 基准上的成功率已提升至约 62%，但精细操作等复杂任务成功率仍不足 15%，表明具身智能仍处于从语义理解向复杂操作能力突破的关键阶段。同时，RoboChallenge 已吸引来自中国、美国、新加坡、日本等多个国家和地区的开发者参与，开源模型、开放数据和开放评测正在成为推动技术创新的重要力量。 + +在具身智能时代，开源正在从软件共享走向更广泛的开放协作。未来需要共同推动模型、数据、机器人接口、评测基准和开放标准的发展。RoboChallenge 致力于连接全球开发者、机器人厂商、科研机构和产业伙伴，建设开放、可信、可复现的真机测评生态，让更多开发者共享真实机器人资源，加速具身智能创新。 + +### 讲师: + + +
+ +Emily Chen: Co-Founder of Kaiyuanshe + +陈阳（Emily Chen），开源社联合创始人，现任开源社理事会成员。她是深耕开源社区多年的实践者、组织者与连接者，致力于推动开源文化在全球与中国的发展与融合。 + +2008年，陈阳创立了 GNOME.Asia 社区，将 GNOME.Asia 峰会成功带到十多个亚洲国家和地区。2010年，她担任 GNOME 基金会董事会成员，并于2014年在瑞典 GUADEC 大会上荣获 GNOME 基金会最高荣誉——“GNOME Pants Award”。 + +自2017年起，陈阳多次组织 GitHub CEO、COO 及全球开源领袖的访华之行，积极推动国际平台与中国开源生态的交流与合作。她曾任OSI 中国联盟代表、中国开源促进联盟副秘书长（2016），在国际标准组织沟通方面发挥了重要作用。 + +她发起了多个在中国极具影响力的开源项目，包括开源社年度报告、中国开源年会（COSCon）以及“33位开源先行者”榜单。此外，她还是 Mozilla 的核心贡献者、Google Summer of Code（GSoC）导师，以及2024 开源议会（Open Source Congress）的组织者之一。现任开放原子开源基金会 TOC 导师。 + +作为中国开源领域的见证者与推动者，陈阳持续以开放、协作、连接的姿态，探索 AI 时代下全球开源的新范式。 \ No newline at end of file diff --git a/content/sessions/keynote-1300492.md b/content/sessions/keynote-1300492.md new file mode 100644 index 000000000..7e8cff3c3 --- /dev/null +++ b/content/sessions/keynote-1300492.md @@ -0,0 +1,22 @@ +--- +title: "Opening Talk: Open Source in the Agentic Era" +date: "2026-08-07T09:00:00" +track: "keynote" +presenters: "Willem Jiang" +stype: "Chinese Session" +room: "MainRoom - YiHe Hall" +--- + +AI is rapidly reshaping how software is created. Code generation, testing, documentation, and review are becoming more efficient, yet the challenges of understanding code, maintaining it over time, and sharing responsibility do not disappear. When code grows faster than human understanding, open-source communities may also accumulate a new form of cognitive debt. + +In these opening remarks, Willem Jiang draws on his twenty-year journey in the Apache community. He will explore the opportunities and trust challenges facing open source in the agentic era and invite a new generation of developers to help build better ways for humans and AI to collaborate. + +### Speakers: + + +
+ +Willem Jiang: Independent Open Source and Agentic AI Consultant + +Willem Jiang (Jiang Ning) is an independent open source and Agentic AI consultant and a maintainer of DeerFlow. He previously served as an Open Source Evangelist at ByteDance’s OSPO and as a member of the Board of Directors of the Apache Software Foundation. +Willem has contributed to numerous Apache projects, including Apache Camel, CXF, ServiceMix, and ServiceComb. Before joining ByteDance, he spent six years at Huawei as an open source technical expert. Earlier in his career, he was a Principal Engineer at Red Hat, where he worked on Fuse ESB, and held engineering roles at FuseSource, IONA, and TravelSky. \ No newline at end of file diff --git a/content/sessions/keynote-1300492.zh.md b/content/sessions/keynote-1300492.zh.md new file mode 100644 index 000000000..16e4dec28 --- /dev/null +++ b/content/sessions/keynote-1300492.zh.md @@ -0,0 +1,21 @@ +--- +title: "开场演讲：Agentic 时代的开源" +date: "2026-08-07T09:00:00" +track: "keynote" +presenters: "Willem Jiang" +stype: "中文演讲" +room: "主会场 - 颐和厅" +--- + +AI 正在迅速重塑软件的创造方式。代码生成、测试、文档与评审正变得愈发高效，但理解代码、长期维护代码以及分担责任这些挑战并未随之消失。当代码增长的速度超过人类的理解速度时，开源社区也可能积累起一种新的认知债务。 + +在本场开场演讲中，Willem Jiang 将回顾自己在 Apache 社区二十年的历程。他将探讨 Agentic 时代开源所面临的机遇与信任挑战，并邀请新一代开发者共同构建人与 AI 更好的协作方式。 + +### 讲师: + + +
+ +Willem Jiang: 独立开源与 Agentic AI 顾问 + +Willem Jiang（姜宁）是一名独立的开源与 Agentic AI 顾问，也是 DeerFlow 的维护者。他此前曾任字节跳动 OSPO 的开源布道师，以及 Apache 软件基金会董事会成员。Willem 曾为众多 Apache 项目做出贡献，包括 Apache Camel、CXF、ServiceMix 和 ServiceComb。在加入字节跳动之前，他在华为担任开源技术专家六年。在职业生涯早期，他曾任 Red Hat 首席工程师，参与 Fuse ESB 的开发，并曾在 FuseSource、IONA 和 TravelSky 担任工程职务。 \ No newline at end of file diff --git a/content/sessions/keynote-1301244.md b/content/sessions/keynote-1301244.md new file mode 100644 index 000000000..8c3e40e80 --- /dev/null +++ b/content/sessions/keynote-1301244.md @@ -0,0 +1,21 @@ +--- +title: "When Code Is No Longer Scarce: Capital, Open Source & SourceDance" +date: "2026-08-09T10:00:00" +track: "keynote" +presenters: "Lucy Chen" +stype: "Chinese Session" +room: "MainRoom - YiHe Hall" +--- + +As AI makes code generation increasingly abundant and inexpensive, what remains truly scarce in open source? This talk reexamines the value of open source in the AI era from the perspective of capital. In 2026, funding is more abundant, yet increasingly concentrated, while the defensibility of open source projects is shifting from distribution efficiency toward trusted adoption, maintainer resilience, and community governance. +Drawing on SourceDance—an AI-enabled open source internship initiative—the talk explores how lower barriers to participation can be translated into a stronger pipeline of high-quality contributors. It also offers a framework for investors, entrepreneurs, and maintainers to assess the long-term value of an open source project. + + +### Speakers: + + +
+ +Lucy Chen: Singapore VC, EIR + +Lucy Chen is an EIR at one of VC Capital (Singapore, $2B+ AUM), focused on AI infrastructure deal sourcing through open-source signals. She created the OSS Investment Scorecard — the first open-source VC evaluation framework for AI projects — which reached 100K+ reads in its first week with contributions from 6+ USD-denominated funds. She advises SourceDance.AI on project selection, connecting open-source community health metrics with venture capital decision-making in the global AI ecosystem. \ No newline at end of file diff --git a/content/sessions/keynote-1301244.zh.md b/content/sessions/keynote-1301244.zh.md new file mode 100644 index 000000000..70eac5cd8 --- /dev/null +++ b/content/sessions/keynote-1301244.zh.md @@ -0,0 +1,19 @@ +--- +title: "当代码不再稀缺：资本、开源与一场 SourceDance" +date: "2026-08-09T10:00:00" +track: "keynote" +presenters: "Lucy Chen" +stype: "中文演讲" +room: "主会场 - 颐和厅" +--- + +当 AI 让代码生成趋近于廉价，开源项目真正稀缺的还会是代码吗？本次演讲从资本视角重新审视 AI 时代的开源价值：2026 年的资金更充裕，也更集中；开源的护城河正从“分发效率”转向可信采用、维护者韧性与社区治理。结合 AI 开源实习项目 SourceDance 的探索，分享如何把更低的参与门槛转化为更高质量的贡献者供给，并讨论投资人、创业者与维护者应如何衡量一个开源项目的长期价值。 + +### 讲师: + + +
+ +Lucy Chen: Singapore VC, EIR + +Lucy Chen 是新加坡某 VC 机构（管理规模超 20 亿美元）的 EIR（入驻企业家），专注于通过开源信号进行 AI 基础设施项目的投资寻源。她创建了 OSS Investment Scorecard--首个面向 AI 项目的开源 VC 评估框架--发布首周阅读量即突破 10 万，并获得 6 家以上美元基金参与共建。她为 SourceDance.AI 提供项目遴选咨询，在全球 AI 生态中将开源社区健康度指标与风险投资决策相连接。 \ No newline at end of file diff --git a/content/sessions/keynote-1301331.md b/content/sessions/keynote-1301331.md new file mode 100644 index 000000000..c1f7f55b6 --- /dev/null +++ b/content/sessions/keynote-1301331.md @@ -0,0 +1,30 @@ +--- +title: "Building the Open Embodied AI Ecosystem:Infrastructure, Data, Models, and Paths to Commercialization" +date: "2026-08-09T10:40:00" +track: "keynote" +presenters: "Jiaqi Sun" +stype: "Chinese Session" +room: "MainRoom - YiHe Hall" +--- + +Embodied AI is bringing artificial intelligence from the digital world into the physical world. Yet its development continues to face significant challenges, including closed software ecosystems, high-cost data, and fragmented toolchains. This keynote will explore how to build an open, portable, and commercially sustainable embodied AI ecosystem across three foundational layers: infrastructure, data, and models. + +Drawing on Motphys’s experience in developing simulation and training platforms, open simulation benchmarks and evaluation frameworks, as well as leading open academic collaborations such as GS-Playground and UniLab, the talk will examine how open co-development and cross-platform collaboration can reduce dependence on any single hardware architecture or proprietary software ecosystem. The goal is to enable simulation, training, and inference workloads to run more flexibly across platforms from NVIDIA, AMD, Intel, and a broader range of European and Chinese computing hardware providers. + +The keynote will also discuss how simulation infrastructure can serve as the foundation for integrating internet-scale video, human demonstrations, real-robot data, and synthetic simulation data into a multimodal data hub for embodied AI training. Such an infrastructure can support the continuous accumulation, reuse, and closed-loop iteration of data across different tasks, models, and robotic embodiments. + +As generative AI reshapes the software industry and challenges traditional open-source models, the talk will revisit the boundaries of openness: which foundational capabilities should be openly shared, and which data assets, models, and industry-specific capabilities should remain sources of commercial differentiation. Open source must establish a new balance among technological autonomy, ecosystem collaboration, commercial returns, and long-term maintenance. + +Scaling embodied AI requires an open ecosystem that connects infrastructure, data, models, computing hardware providers, robotics companies, research institutions, and developers. + + +### Speakers: + + +
+ +Jiaqi Sun: Motphys, Co-founder (谋先飞联合创始人) + +Jiaqi Sun, co-founder of Motphys, PhD from the Department of Mechanics at Peking University. He has extensive research and industry experience in physics simulation, computational geometry and graphics algorithms, heterogeneous CPU/GPU computing, and the integration of artificial intelligence with scientific computing. For more than a decade, he has worked at the forefront of industrial physics solver development, leading the creation of integrated software that connects geometric processing with multiphysics simulation. The resulting technologies and products have been applied across automotive engineering, aerospace, data centers, renewable energy, semiconductor manufacturing, and other industrial simulation domains. + +In recent years, Jiaqi has focused on integrating physics simulation with embodied AI at Motphys. He has driven the development of the Motrix family of simulation and training platforms and the MotrixBench open simulation benchmarking framework, while also supporting academic collaborations such as GS-Playground and UniLab. His work explores open infrastructure for Real2Sim2Real pipelines, robot learning, and heterogeneous computing platforms. He also studies how to build an open-source embodied AI ecosystem that balances technological independence, collaborative innovation, and long-term commercial sustainability. \ No newline at end of file diff --git a/content/sessions/keynote-1301331.zh.md b/content/sessions/keynote-1301331.zh.md new file mode 100644 index 000000000..3fdeb523c --- /dev/null +++ b/content/sessions/keynote-1301331.zh.md @@ -0,0 +1,26 @@ +--- +title: "构建具身智能开放生态：基础设施、数据、模型与商业化路径" +date: "2026-08-09T10:40:00" +track: "keynote" +presenters: "Jiaqi Sun" +stype: "中文演讲" +room: "主会场 - 颐和厅" +--- + +具身智能正在推动 AI 从数字世界走向真实物理世界，但其发展仍面临封闭的软件生态、高成本数据和碎片化工具链等挑战。本次演讲将围绕基础设施、数据与模型，探讨如何构建开放、可移植且具有商业可持续性的具身智能生态。 + +结合 Motphys 在仿真训练平台、仿真评测体系，以及主导建设GS-Playground、UniLab 等开源学术合作项目的实践经验，讨论如何通过开放共建与跨平台协作，降低具身智能对单一硬件架构和专有软件生态的依赖，使仿真、训练与推理任务能够更灵活地运行在 NVIDIA、AMD、Intel，以及欧洲、中国不同算力供应商的平台之上。 + +演讲将介绍如何以仿真基础设施为核心，融合互联网视频、人类示范、真机数据和仿真合成数据，建设服务于具身训练的多模态数据中枢，推动数据在不同任务、模型和机器人本体之间持续积累、复用与闭环迭代。 + +面对生成式 AI 对软件产业和传统开源模式带来的冲击，演讲还将重新审视开源的边界：哪些基础能力适合开放共享，哪些数据、模型和行业能力需要保留商业差异化。开源需要在技术自主、生态协作、商业回报和长期维护之间建立新的平衡。 +具身智能的规模化发展，需要一个连接基础设施、数据、模型、算力厂商、机器人企业、研究机构与开发者的开放生态。 + +### 讲师: + + +
+ +孙家琦，谋先飞（Motphys）联合创始人，北京大学计算力学博士。长期从事物理仿真、几何与图形学算法、CPU/GPU 异构并行计算，以及 AI 与科学计算融合方向的研究和产业实践。十余年来，他一直深耕工业级物理求解器开发前沿，主导开发了几何处理与多物理场求解一体化软件，相关技术与产品已应用于汽车、航空航天、数据中心、新能源和半导体等工业物理仿真领域。 + +近年来，他在Motphys致力于将物理仿真与具身智能融合，推动Motrix系列仿真训练平台与MotrixBench开放仿真评测体系研发，并支持学术合作项目。他的工作正在建设面向 Real2Sim2Real流程、机器人学习和多元异构算力平台的开放基础设施。同时探索如何构建一个在技术自主、协同创新与长期商业可持续性之间取得平衡的开源具身智能生态。 \ No newline at end of file diff --git a/content/sessions/messaging-1196418.md b/content/sessions/messaging-1196418.md new file mode 100644 index 000000000..ccab7b999 --- /dev/null +++ b/content/sessions/messaging-1196418.md @@ -0,0 +1,21 @@ +--- +title: "Apache RocketMQ Routing Control Plane Evolution" +date: "2026-08-07T16:15:00" +track: "messaging" +presenters: "Xiao Yang" +stype: "Chinese Session" +room: "JingYi Hall" +--- + +This session explains how we evolved a lightweight external routing control plane for Apache RocketMQ to support large-cluster governance without changing brokers, nameservers, or client access patterns. + +As multiple business lines gradually shared one common cluster, the real challenge shifted from throughput to governance: faster failure convergence, controllable topic migration, and flexible lane-based traffic isolation. In this talk, I will show why these problems are better solved in the routing layer, why existing NameServer and Broker mechanisms are not sufficient, and how broker write isolation, topic migration, and lane-based traffic isolation can be unified in one control plane to support large-cluster split with a smaller blast radius and more controllable governance. + +### Speakers: + + +
+ +Xiao Yang: Middleware Engineer at Tongcheng Travel, Apache RocketMQ Committer + +Responsible for the architectural evolution, reliability engineering, and platform development of messaging middleware. \ No newline at end of file diff --git a/content/sessions/messaging-1196418.zh.md b/content/sessions/messaging-1196418.zh.md new file mode 100644 index 000000000..e4b1a3cf8 --- /dev/null +++ b/content/sessions/messaging-1196418.zh.md @@ -0,0 +1,21 @@ +--- +title: "Apache RocketMQ 路由控制面演进" +date: "2026-08-07T16:15:00" +track: "messaging" +presenters: "Xiao Yang" +stype: "中文演讲" +room: "静宜厅" +--- + +本次分享将讲解我们如何为 Apache RocketMQ 演进出一套轻量级的外部路由控制面，在不改动 broker、nameserver 以及客户端访问模式的前提下，支撑大规模集群的治理。 + +随着多条业务线逐步共用同一个集群，真正的挑战从吞吐量转向治理：更快的故障收敛、可控的 topic 迁移，以及灵活的基于泳道（lane）的流量隔离。在本次演讲中，我将说明为什么这些问题更适合在路由层解决，为什么现有的 NameServer 和 Broker 机制并不足够，以及如何将 broker 写隔离、topic 迁移和基于泳道的流量隔离统一在同一个控制面中，从而以更小的爆炸半径、更可控的治理方式支撑大规模集群的拆分。 + +### 讲师: + + +
+ +Xiao Yang：同程旅行中间件工程师，Apache RocketMQ Committer + +负责消息中间件架构演进、稳定性治理与平台化能力建设。 \ No newline at end of file diff --git a/content/sessions/messaging-1203579.md b/content/sessions/messaging-1203579.md new file mode 100644 index 000000000..a6bc1d08a --- /dev/null +++ b/content/sessions/messaging-1203579.md @@ -0,0 +1,63 @@ +--- +title: "Cost Challenges of Apache Kafka in the AI Era: Optimization Practice and Latest Community Progress" +date: "2026-08-07T14:30:00" +track: "messaging" +presenters: "Jian Fu" +stype: "Chinese Session" +room: "JingYi Hall" +--- + +The explosive growth of data driven by AI has significantly increased Kafka workloads and operational costs. This session focuses on how to optimize Kafka costs, covering the following key aspects: + +1. The composition of Kafka costs and their relationship to system architecture +2. How to kick off Kafka cost optimization, e.g., the motivation behind KIP-1227 (exposing RackID) +3. Recent cost-related KIPs and PRs (e.g., KAFKA-19909, KIP-1123, KIP-1241, KIP-1150) and how they help reduce costs +4. Beyond Kafka itself, additional approaches to lowering costs (e.g., non-technical efforts, modifications to KIP-405 tiered storage plugins) +5. Common pitfalls and challenges encountered in real-world cost optimization +6. The role and features of AI in cost optimization + +Target Audience +Engineers, architects, and technical leaders who are concerned with Kafka cost management and optimization practices or those who want to learn Kafka architecture from a cost perspective + +Expected Takeaways +Gain a clear understanding of Kafka cost structure, optimization strategies, and common pitfalls, along with the latest technical advancements and AI-driven practices. These methodologies can also be applied to cost optimization in other software systems. + +### Speakers: + + +
+ +Jian Fu: Zoom-Engineering Manager + +Engineering Manager at Zoom, formerly at Cisco, leading the development and operations of Zoom’s messaging queue systems, with 15+ years of experience building high-performance, highly available, and elastic systems. + +Research: + 1 Patent in USA , 10+ Published paper + +Certification: + Network Engineer + Information Systems Project Manager + +Contributor of open sources: + Netty + Apache kafka + Influxdb/Influxdb-java + Redis/Jedis + Spring/Spring-data-redis + Jenkins + Apache zookeeper + +Author of the books: + authored: + Netty原理剖析与应用(Netty author Trustin Lee recommended） + co-authored: + (微服务之道：度量驱动开发) + (完美测试：软件测试系列最佳实践) + +Author of the Kafka improvement proposals: + Four Kafka improvement proposals (KIP) Accept and implemented + +Author of the online courses: + 极客时间: + (Spring 编程常见错误50例) + (Netty 源码剖析与实战) \ No newline at end of file diff --git a/content/sessions/messaging-1203579.zh.md b/content/sessions/messaging-1203579.zh.md new file mode 100644 index 000000000..cdbcb6be1 --- /dev/null +++ b/content/sessions/messaging-1203579.zh.md @@ -0,0 +1,63 @@ +--- +title: "AI 时代 Apache Kafka 的成本挑战：优化实践与社区最新进展" +date: "2026-08-07T14:30:00" +track: "messaging" +presenters: "Jian Fu" +stype: "中文演讲" +room: "静宜厅" +--- + +AI 驱动的数据爆发式增长，使得 Kafka 的工作负载和运维成本显著上升。本次分享聚焦于如何优化 Kafka 成本，涵盖以下几个关键方面： + +1. Kafka 成本的构成，以及其与系统架构之间的关系 +2. 如何着手 Kafka 成本优化，例如 KIP-1227（暴露 RackID）背后的动机 +3. 近期与成本相关的 KIP 和 PR（例如 KAFKA-19909、KIP-1123、KIP-1241、KIP-1150）以及它们如何帮助降低成本 +4. 除 Kafka 本身之外，其他降低成本的途径（例如非技术层面的努力、对 KIP-405 分层存储插件的改造） +5. 在真实成本优化过程中遇到的常见陷阱与挑战 +6. AI 在成本优化中所扮演的角色及其特性 + +目标听众 +关注 Kafka 成本管理与优化实践的工程师、架构师和技术负责人，或希望从成本视角了解 Kafka 架构的人士 + +预期收获 +清晰地理解 Kafka 的成本结构、优化策略与常见陷阱，以及最新的技术进展和 AI 驱动的实践。这些方法论同样可以应用到其他软件系统的成本优化中。 + +### 讲师: + + +
+ +Jian Fu：Zoom，工程经理 + +Zoom 工程经理，曾就职于 Cisco，负责 Zoom 消息队列系统的开发与运维，拥有 15 年以上构建高性能、高可用、弹性系统的经验。 + +研究： + 1 项美国专利，10 余篇已发表论文 + +认证： + 网络工程师 + 信息系统项目管理师 + +开源贡献者： + Netty + Apache Kafka + InfluxDB/InfluxDB-java + Redis/Jedis + Spring/Spring-data-redis + Jenkins + Apache Zookeeper + +著作： + 独著： + 《Netty 原理剖析与应用》（Netty 作者 Trustin Lee 推荐） + 合著： + 《微服务之道：度量驱动开发》 + 《完美测试：软件测试系列最佳实践》 + +Kafka 改进提案（KIP）作者： + 四项 KIP 已被采纳并实现 + +在线课程作者： + 极客时间： + 《Spring 编程常见错误 50 例》 + 《Netty 源码剖析与实战》 \ No newline at end of file diff --git a/content/sessions/messaging-1209406.md b/content/sessions/messaging-1209406.md new file mode 100644 index 000000000..f80313f78 --- /dev/null +++ b/content/sessions/messaging-1209406.md @@ -0,0 +1,31 @@ +--- +title: "Challenges and Breakthroughs in TDMQ for Apache Pulsar: Multi-Modal Storage Delayed Messaging" +date: "2026-08-07T15:00:00" +track: "messaging" +presenters: "Xiaolong Ran, ZhenJiang Wang" +stype: "Chinese Session" +room: "JingYi Hall" +--- + +In the large-scale deployment of Apache Pulsar, storage architecture and delayed messages are two unavoidable core challenges. + +On the delayed messaging side, the community solution suffers from message hole issues, high index memory consumption, and long index loading times when dealing with a large volume of delayed messages with randomly distributed delay times. While scaling partitions, increasing bandwidth, and expanding memory can partially mitigate these issues, the cost is prohibitively high and difficult to standardize as a cloud product. This talk will share how TDMQ for Apache Pulsar addresses these challenges, enabling stable large-scale delayed messaging while significantly reducing resource costs. + +On the delayed messaging side, the community solution faces several challenges when handling large volumes of delayed messages with highly random delay time distributions, including excessive memory consumption of subscription progress objects and prolonged loading times for delayed message indexes. Although these issues can be partially mitigated by expanding partitions, increasing bandwidth, and scaling up memory, the cost is prohibitively high from an economic perspective, and it is difficult to flexibly adapt to diverse business scenarios from a cloud product standardization standpoint. This talk will share TDMQ for Apache Pulsar's optimization approach and production practices to address these challenges, ensuring stable operation of large-scale delayed messaging while significantly reducing resource costs. + +Both topics are based on real production experience from the TDMQ team, aiming to provide valuable insights for large-scale Pulsar deployments on the cloud. + +### Speakers: + + +
+ +Xiaolong Ran: Senior R&D Engineer at Tencent Cloud Middleware + +Apache Pulsar Committer, author of Pulsar Go Functions, Go Client, and pulsarctl, as well as a Maintainer and one of the core developers of RoP. + +
+ +ZhenJiang Wang: Senior R&D Engineer at Tencent Cloud Middleware + +Apache Pulsar contributor, Open-source community enthusiasts \ No newline at end of file diff --git a/content/sessions/messaging-1209406.zh.md b/content/sessions/messaging-1209406.zh.md new file mode 100644 index 000000000..5f81af095 --- /dev/null +++ b/content/sessions/messaging-1209406.zh.md @@ -0,0 +1,31 @@ +--- +title: "TDMQ for Apache Pulsar 的挑战与突破：多模态存储与延迟消息" +date: "2026-08-07T15:00:00" +track: "messaging" +presenters: "Xiaolong Ran, ZhenJiang Wang" +stype: "中文演讲" +room: "静宜厅" +--- + +在 Apache Pulsar 的大规模部署中，存储架构与延迟消息是两个绕不开的核心挑战。 + +在延迟消息方面，社区方案在处理大量、延迟时间随机分布的延迟消息时，会遇到消息空洞（message hole）、索引内存占用过高以及索引加载时间过长等问题。虽然通过扩分区、加带宽、扩内存可以在一定程度上缓解这些问题，但其成本高得令人难以承受，也很难标准化为云产品。本次演讲将分享 TDMQ for Apache Pulsar 如何应对这些挑战，在实现稳定的大规模延迟消息的同时显著降低资源成本。 + +在延迟消息方面，社区方案在处理大量、延迟时间高度随机分布的延迟消息时会面临若干挑战，包括订阅进度对象内存占用过高、延迟消息索引加载时间过长等。尽管通过扩分区、加带宽、扩内存可在一定程度上缓解这些问题，但从经济角度看成本高得难以承受，从云产品标准化的角度也难以灵活适配多样的业务场景。本次演讲将分享 TDMQ for Apache Pulsar 应对这些挑战的优化思路与生产实践，保障大规模延迟消息的稳定运行，同时显著降低资源成本。 + +这两个议题都基于 TDMQ 团队的真实生产经验，旨在为云端大规模 Pulsar 部署提供有价值的参考。 + +### 讲师: + + +
+ +Xiaolong Ran：腾讯云中间件高级研发工程师 + +Apache Pulsar Committer，Pulsar Go Functions、Go Client 和 pulsarctl 的作者，同时也是 RoP 的 Maintainer 和核心开发者之一。 + +
+ +ZhenJiang Wang: 腾讯云中间件高级研发工程师 + +Apache Pulsar contributor, 开源社区爱好者 \ No newline at end of file diff --git a/content/sessions/messaging-1211234.md b/content/sessions/messaging-1211234.md new file mode 100644 index 000000000..96aa24d99 --- /dev/null +++ b/content/sessions/messaging-1211234.md @@ -0,0 +1,25 @@ +--- +title: "Rethinking High Availability for Apache RocketMQ on the Cloud: Protocol-Fenced Takeover for Second-L" +date: "2026-08-07T15:45:00" +track: "messaging" +presenters: "Rongtong Jin" +stype: "Chinese Session" +room: "JingYi Hall" +--- + +High availability for stateful messaging systems in the cloud is often constrained by a difficult trade-off among cost, steady-state performance, and failover speed. Replication-based approaches can reduce recovery time, but they usually introduce extra write amplification, resource overhead, and operational complexity. In contrast, single-replica deployments preserve performance and cost efficiency, yet their failover path in Kubernetes is often dominated by slow and unpredictable volume detach/attach workflows. + +In this session, I will share how we rethought high availability for Apache RocketMQ on the cloud and built a production-proven architecture that achieves second-level failover without adding extra data replicas in the steady state. The key idea is to keep RocketMQ’s normal storage path unchanged while redesigning the takeover path around Multi-Attach block storage and NVMe Persistent Reservation. This protocol-fenced takeover model allows a new node to safely preempt disk ownership, prevent split-brain, and resume service quickly without waiting for the traditional storage migration process. + +I will walk through the architecture, failure detection and takeover workflow, safety and consistency considerations, and the implementation details in Apache RocketMQ. The talk will also cover how decentralized lease probing, node-side takeover orchestration, and crash-consistent recovery work together to shorten and stabilize the failover critical path. + +Finally, I will share production experience from Alibaba Cloud, where this architecture has already been deployed for Apache RocketMQ services and validated through large-scale disaster recovery drills. The underlying work behind this session has also been accepted by the FSE 2026 Industry track, while this talk will focus on the production architecture, engineering trade-offs, and practical lessons learned from Apache RocketMQ. Attendees will gain practical insights into designing highly available stateful messaging systems in cloud-native environments, and learn how to balance recovery objectives with cost and performance in real-world production systems. + +### Speakers: + + +
+ +Rongtong Jin: Technical Expert at Alibaba Cloud, Apache RocketMQ PMC Member + +Rongtong Jin is a Technical Expert at Alibaba Cloud and a PMC Member of Apache RocketMQ. He works on the architecture and production operations of cloud messaging systems, with a focus on high availability, storage reliability, and large-scale stateful services. He has extensive experience in the cloud evolution, production rollout, and stability engineering of Apache RocketMQ. \ No newline at end of file diff --git a/content/sessions/messaging-1211234.zh.md b/content/sessions/messaging-1211234.zh.md new file mode 100644 index 000000000..3fdc4c8af --- /dev/null +++ b/content/sessions/messaging-1211234.zh.md @@ -0,0 +1,25 @@ +--- +title: "重新思考云端 Apache RocketMQ 的高可用：面向秒级故障恢复的 Protocol-Fenced 接管" +date: "2026-08-07T15:45:00" +track: "messaging" +presenters: "Rongtong Jin" +stype: "中文演讲" +room: "静宜厅" +--- + +云端有状态消息系统的高可用，往往受制于成本、稳态性能与故障恢复速度之间艰难的权衡。基于副本（replication）的方案可以缩短恢复时间，但通常会带来额外的写放大、资源开销和运维复杂度。相比之下，单副本部署能够保持性能与成本效率，但其在 Kubernetes 中的故障恢复路径往往被缓慢且不可控的 volume detach/attach 流程所主导。 + +在本次演讲中，我将分享我们如何重新思考云端 Apache RocketMQ 的高可用，并构建了一套经过生产验证的架构——在稳态下不增加额外数据副本即可实现秒级故障恢复。核心思路是保持 RocketMQ 常规存储路径不变，转而围绕 Multi-Attach 块存储和 NVMe Persistent Reservation 重新设计接管路径。这种 protocol-fenced 接管模型让新节点能够安全地抢占磁盘所有权、防止脑裂，并快速恢复服务，而无需等待传统的存储迁移流程。 + +我将讲解整体架构、故障检测与接管流程、安全性与一致性考量，以及在 Apache RocketMQ 中的实现细节。演讲还会介绍去中心化的租约探测（lease probing）、节点侧的接管编排，以及崩溃一致性恢复（crash-consistent recovery）如何协同工作，从而缩短并稳定故障恢复的关键路径。 + +最后，我将分享来自阿里云的生产经验——这套架构已部署于 Apache RocketMQ 服务，并通过了大规模容灾演练的验证。本议题背后的底层工作也被 FSE 2026 Industry track 收录，而本次演讲将聚焦于生产架构、工程权衡，以及从 Apache RocketMQ 中总结的实践教训。听众将获得在云原生环境下设计高可用有状态消息系统的实用洞察，并学会在真实生产系统中如何在恢复目标与成本、性能之间取得平衡。 + +### 讲师: + + +
+ +Rongtong Jin：阿里云技术专家，Apache RocketMQ PMC 成员 + +Rongtong Jin 是阿里云的技术专家，也是 Apache RocketMQ 的 PMC 成员。他从事云端消息系统的架构设计与生产运维，专注于高可用、存储可靠性和大规模有状态服务。他在 Apache RocketMQ 的云化演进、生产落地和稳定性工程方面拥有丰富经验。 \ No newline at end of file diff --git a/content/sessions/messaging-1211843.md b/content/sessions/messaging-1211843.md new file mode 100644 index 000000000..b55b26473 --- /dev/null +++ b/content/sessions/messaging-1211843.md @@ -0,0 +1,19 @@ +--- +title: "Xiaomi RocketMQ Production Practices in Large-Scale Clusters" +date: "2026-08-09T15:15:00" +track: "messaging" +presenters: "Kunming Han" +stype: "Chinese Session" +room: "JingYi Hall" +--- + +This sharing focuses on Xiaomi’s practical experience in applying RocketMQ and MQTT in large-scale cluster scenarios, with an emphasis on the core challenges faced by Xiaomi in the field of message middleware and the corresponding solutions.It covers the optimization of ordered messages, disaster recovery and current limiting mechanisms, and multi-dimensional performance tuning for RocketMQ, as well as stability assurance and architectural optimization practices of MQTT in scenarios with tens of millions of connections.The session aims to provide the industry with reference insights for the implementation of large-scale messaging systems. + +### Speakers: + + +
+ +Kunming Han: Xiaomi, Software R&D Engineer + +Mainly responsible for message queue related work in Xiaomi \ No newline at end of file diff --git a/content/sessions/messaging-1211843.zh.md b/content/sessions/messaging-1211843.zh.md new file mode 100644 index 000000000..a91a1924a --- /dev/null +++ b/content/sessions/messaging-1211843.zh.md @@ -0,0 +1,19 @@ +--- +title: "小米 RocketMQ 在大规模集群中的生产实践" +date: "2026-08-09T15:15:00" +track: "messaging" +presenters: "Kunming Han" +stype: "中文演讲" +room: "静宜厅" +--- + +本次分享聚焦于小米在大规模集群场景下应用 RocketMQ 与 MQTT 的实践经验，重点介绍小米在消息中间件领域所面临的核心挑战及相应的解决方案。内容涵盖 RocketMQ 的顺序消息优化、容灾与限流机制、多维度性能调优，以及 MQTT 在千万级连接场景下的稳定性保障与架构优化实践。本次分享旨在为业界大规模消息系统的落地提供参考与借鉴。 + +### 讲师: + + +
+ +Kunming Han：小米，软件研发工程师 + +主要负责小米的消息队列相关工作 \ No newline at end of file diff --git a/content/sessions/messaging-1212296.md b/content/sessions/messaging-1212296.md new file mode 100644 index 000000000..ad966c3e5 --- /dev/null +++ b/content/sessions/messaging-1212296.md @@ -0,0 +1,26 @@ +--- +title: "AI Link，AI access layer of Tencent Cloud MQTT" +date: "2026-08-07T16:45:00" +track: "messaging" +presenters: "Wei Yang, Senze Zhang" +stype: "Chinese Session" +room: "JingYi Hall" +--- + +AI Link is the AI access layer of Tencent Cloud MQTT — enabling AI agents to discover, understand, and control physical devices. Built on top of the MQTT broker, it exposes physical devices as MCP-compatible tools, aggregates real-time state and historical context for AI reasoning + +### Speakers: + + +
+ +Wei Yang: Tencent Cloud TDMQ Product Manager + +Wei Yang, Senior Product Manager of Cloud Middleware at Tencent Cloud, leading the commercialization of Tencent Cloud MQTT. + + +
+ +Senze Zhang: Apache RocketMQ Committer + +Senze Zhang is a seasoned engineer specializing in the field of message queuing systems. With extensive experience in the design, implementation, and optimization of high-performance messaging solutions, Senze has participated in deploying various industry-leading message queuing platforms including RocketMQ, Kafka, and MQTT. \ No newline at end of file diff --git a/content/sessions/messaging-1212296.zh.md b/content/sessions/messaging-1212296.zh.md new file mode 100644 index 000000000..0cb948ed1 --- /dev/null +++ b/content/sessions/messaging-1212296.zh.md @@ -0,0 +1,26 @@ +--- +title: "AI Link：腾讯云 MQTT 的 AI 接入层" +date: "2026-08-07T16:45:00" +track: "messaging" +presenters: "Wei Yang, Senze Zhang" +stype: "中文演讲" +room: "静宜厅" +--- + +AI Link 是腾讯云 MQTT 的 AI 接入层——让 AI agent 能够发现、理解并控制物理设备。它构建于 MQTT broker 之上，将物理设备暴露为 MCP 兼容的工具，并聚合实时状态与历史上下文以供 AI 推理使用。 + +### 讲师: + + +
+ +Wei Yang：腾讯云 TDMQ 产品经理 + +Wei Yang，腾讯云云中间件高级产品经理，主导腾讯云 MQTT 的商业化落地。 + + +
+ +Senze Zhang：Apache RocketMQ Committer + +Senze Zhang 是一位资深的工程师，专注于消息队列系统领域。他在高性能消息解决方案的设计、实现与优化方面经验丰富，曾参与部署包括 RocketMQ、Kafka、MQTT 在内的多种业界领先的消息队列平台。 \ No newline at end of file diff --git a/content/sessions/messaging-1212333.md b/content/sessions/messaging-1212333.md new file mode 100644 index 000000000..e1124cc7d --- /dev/null +++ b/content/sessions/messaging-1212333.md @@ -0,0 +1,23 @@ +--- +title: "Apache Pulsar multi-cluster disaster recovery and progressive traffic migration" +date: "2026-08-09T15:45:00" +track: "messaging" +presenters: "Dezhi LIU" +stype: "Chinese Session" +room: "JingYi Hall" +--- + +Multi-cluster Pulsar operations face two core challenges: rapid disaster recovery failover in the event of a primary cluster failure, and smooth traffic migration during cluster relocation. This presentation introduces the design and implementation of a proxy, which unifies disaster recovery and scheduling through three key technologies—directed connection closure, native weighted routing, and a progressive migration engine—to achieve arbitrary-precision traffic allocation and zero-explosion-radius failover. + +### Speakers: + + +
+ +Dezhi LIU: AscentStream Technology co-founder + +● Co-founder of AscentStream Technology + +● 10+ years of experience in large-scale internet/financial infrastructure development + +● Former expert engineer at Tencent, successfully driving the large-scale deployment of Pulsar on Tencent Group's unified billing platform's message bus (capable of handling hundreds of billions of messages) and Tencent Cloud's financial-grade messaging service (TDMQ). \ No newline at end of file diff --git a/content/sessions/messaging-1212333.zh.md b/content/sessions/messaging-1212333.zh.md new file mode 100644 index 000000000..3eb27024d --- /dev/null +++ b/content/sessions/messaging-1212333.zh.md @@ -0,0 +1,23 @@ +--- +title: "Apache Pulsar 多集群容灾与渐进式流量迁移" +date: "2026-08-09T15:45:00" +track: "messaging" +presenters: "Dezhi LIU" +stype: "中文演讲" +room: "静宜厅" +--- + +多集群 Pulsar 的运维面临两大核心挑战：主集群故障时的快速容灾切换，以及集群搬迁期间的平滑流量迁移。本次分享介绍一款 proxy 的设计与实现，它通过三项关键技术——定向连接关闭、原生加权路由以及渐进式迁移引擎——将容灾与调度统一起来，从而实现任意精度的流量分配与零爆炸半径的故障切换。 + +### 讲师: + + +
+ +Dezhi LIU：Ascentstream Technology 联合创始人 + +● 谙流科技联合创始人 + +● 10 年以上大规模互联网/金融基础架构开发经验 + +● 前腾讯专家工程师，曾成功推动 Pulsar 在腾讯集团统一计费平台消息总线（可处理千亿级消息）以及腾讯云金融级消息服务（TDMQ）上的大规模落地 \ No newline at end of file diff --git a/content/sessions/messaging-1212359.md b/content/sessions/messaging-1212359.md new file mode 100644 index 000000000..68b4a3ff7 --- /dev/null +++ b/content/sessions/messaging-1212359.md @@ -0,0 +1,27 @@ +--- +title: "Real-Time Lakehouse Pipeline: Native Lake Ingestion and Embedded Computation in Disaggregated Kafka" +date: "2026-08-09T13:30:00" +track: "messaging" +presenters: "Hongjian Fei" +stype: "Chinese Session" +room: "JingYi Hall" +--- + +Apache Kafka has established itself as the central nervous system of modern real-time data architectures. Yet in conventional lakehouse designs, data from Kafka typically must traverse external compute engines such as Flink or Spark before it can be landed, modeled, and organized into analytical layers. This multi-hop "messaging → computation → storage" pipeline not only inflates operational complexity but also introduces weak state consistency, elevated end-to-end latency, and a fragmented hot/cold data lifecycle. + +This talk focuses on disaggregated storage-compute architectures and explores the evolution of Kafka's native capabilities in lakehouse scenarios. We aim to restructure the real-time data pipeline through the following core techniques: + +Disaggregated Storage-Compute with Intelligent Data Tiering. Built on a disaggregated storage-compute model with intelligent tiering and a lightweight high-availability architecture, we employ a dual-path Table-Topic synchronization mechanism that strikes an optimal balance between data freshness and system stability. By internalizing ingestion progress into the leader's metadata, we completely eliminate the dependency on external key-value stores, dramatically simplifying the pipeline while preserving strong consistency guarantees. + +Dual Row-Columnar Engine with Strong Consistency Guarantees. Leveraging a pluggable architecture, we achieve protocol-aware parsing coupled with zero-copy direct writes in Apache Parquet columnar format — effectively turning streaming writes into immediate lakehouse persistence. A lightweight distributed transaction coordinator, integrated with an OSS-backed Table Catalog, strictly enforces Exactly-Once semantics and atomic file commits across node failures, delivering high-availability disaster recovery and autonomous state healing for the ingestion pipeline. + +Embedded Declarative Stream Processing. Apache Calcite is embedded as a lightweight operator engine, enabling users to express computation logic through a SQL dialect. With operator pushdown and dynamic compilation of logical plans, filtering, aggregation, and transformation operations are executed in-situ within the Kafka pipeline, right at the data's point of origin. + +### Speakers: + + +
+ +Hongjian Fei: Alibaba Cloud, Middleware Development Engineer, Staff Engineer + +Hongjian Fei (费红健) is an Apache Software Committer who has long been dedicated to development on RocketMQ and Kafka. At Alibaba Cloud, he is one of the core designers of disaggregated storage and compute architecture for Kafka, bringing extensive experience in Kafka development and operations. \ No newline at end of file diff --git a/content/sessions/messaging-1212359.zh.md b/content/sessions/messaging-1212359.zh.md new file mode 100644 index 000000000..6a55ecf85 --- /dev/null +++ b/content/sessions/messaging-1212359.zh.md @@ -0,0 +1,27 @@ +--- +title: "实时湖仓管道：分离式 Kafka 中的原生入湖与内嵌计算" +date: "2026-08-09T13:30:00" +track: "messaging" +presenters: "Hongjian Fei" +stype: "中文演讲" +room: "静宜厅" +--- + +Apache Kafka 已经成为现代实时数据架构的中枢神经。然而在传统的湖仓设计中，来自 Kafka 的数据通常必须先经过 Flink 或 Spark 等外部计算引擎，才能被落库、建模并组织成分析层。这种多跳的"消息 → 计算 → 存储"管道不仅推高了运维复杂度，还带来了状态一致性较弱、端到端时延升高、冷热数据生命周期割裂等问题。 + +本次演讲聚焦于存储计算分离架构，探讨 Kafka 原生能力在湖仓场景下的演进。我们希望通过以下核心技术来重塑实时数据管道： + +存储计算分离与智能数据分层。建立在存储计算分离模型之上，结合智能分层与轻量级高可用架构，我们采用了双路径 Table-Topic 同步机制，在数据新鲜度与系统稳定性之间取得最佳平衡。通过将摄入（ingestion）进度内化到 leader 的元数据中，我们彻底消除对外部 KV 存储的依赖，在保持强一致性保证的同时极大简化了管道。 + +具备强一致性保证的行列双引擎。借助可插拔架构，我们实现了协议感知（protocol-aware）的解析，并结合 Apache Parquet 列式格式的零拷贝直写——有效地把流式写入转化为即时的湖仓持久化。一个轻量级的分布式事务协调器与基于 OSS 的 Table Catalog 相结合，在节点故障情况下严格保证 Exactly-Once 语义与原子文件提交，为摄入管道提供高可用容灾与自主状态自愈能力。 + +内嵌的声明式流处理。我们将 Apache Calcite 作为轻量级算子引擎内嵌进来，让用户可以通过一种 SQL 方言来表达计算逻辑。借助算子下推与逻辑计划的动态编译，过滤、聚合和转换操作可直接在 Kafka 管道中、在数据的产生处就地执行。 + +### 讲师: + + +
+ +Hongjian Fei：阿里云，中间件开发工程师，Staff Engineer + +Hongjian Fei（费红健）是一位 Apache 软件基金会 Committer，长期致力于 RocketMQ 与 Kafka 的开发。在阿里云，他是 Kafka 存储计算分离架构的核心设计者之一，在 Kafka 的开发与运维方面经验丰富。 \ No newline at end of file diff --git a/content/sessions/messaging-1212360.md b/content/sessions/messaging-1212360.md new file mode 100644 index 000000000..15e53af12 --- /dev/null +++ b/content/sessions/messaging-1212360.md @@ -0,0 +1,25 @@ +--- +title: "RocketMQ 5 in Production at Meituan" +date: "2026-08-09T14:00:00" +track: "messaging" +presenters: "Jiangge Zhang" +stype: "Chinese Session" +room: "JingYi Hall" +--- + +Meituan operates one of China's largest internal messaging platforms, originally built on an early Kafka fork—internally known as Mafka. Mafka served us well, but over time hit two hard limits. First, consumer scalability was capped by partition count: adding more consumer instances beyond the number of partitions yielded no throughput gain. Second, the client became a heavyweight artifact—carrying complex routing logic and a wire protocol that made cross-language support a maintenance nightmare. + +Apache RocketMQ 5 addressed both. The Pop consumption model manages message lifecycle per message rather than per partition, allowing a queue to be shared across consumers simultaneously—unlocking horizontal elasticity. The Proxy architecture, built on gRPC, enables lightweight, language-agnostic clients. But RocketMQ 5 out of the box could not cover what Meituan needs at scale: a unified global metadata layer and enterprise-grade traffic isolation across multiple deployment environments and business units. + +This talk walks through the decisions and trade-offs behind our production adoption of RocketMQ 5 at Meituan. We cover three areas: first, how we designed a three-tier topology—tenant, placement group, and cluster—where placement groups serve as blast-radius boundaries within a tenant; second, how we implemented multiple modes of traffic isolation and routing entirely outside RocketMQ, so the broker stays a pure messaging primitive and isolation policies can evolve independently; third, how we built a single authoritative metadata service to keep all components in sync. + +We will ground the talk in real-world results from active production migration. Running at scale has validated our core architectural choices, but also surfaced hard problems we are still working through—from data-plane constraints in RocketMQ 5's replication model to capacity isolation challenges inherent in its shared commit log design. We hope these production learnings are useful to the broader RocketMQ community. + +### Speakers: + + +
+ +Jiangge Zhang: Software Architect at Meituan + +Jiangge Zhang is an architect at Meituan, working on the messaging infrastructure team. He previously worked on SOA middleware and has been focused on large-scale distributed systems throughout his career. \ No newline at end of file diff --git a/content/sessions/messaging-1212360.zh.md b/content/sessions/messaging-1212360.zh.md new file mode 100644 index 000000000..ac2679a7e --- /dev/null +++ b/content/sessions/messaging-1212360.zh.md @@ -0,0 +1,25 @@ +--- +title: "RocketMQ 5 在美团的生产实践" +date: "2026-08-09T14:00:00" +track: "messaging" +presenters: "Jiangge Zhang" +stype: "中文演讲" +room: "静宜厅" +--- + +美团运营着中国最大的内部消息平台之一，它最初构建于一个早期 Kafka 分支之上——内部称为 Mafka。Mafka 曾很好地服务于我们，但随着时间推移，碰到了两个硬性瓶颈。其一，消费者的可扩展性受限于分区数量：当消费者实例数量超过分区数时，吞吐量不会有任何提升。其二，客户端变成了一个沉重的产物——携带复杂的路由逻辑和一套二进制通信协议（wire protocol），使得跨语言支持沦为运维噩梦。 + +Apache RocketMQ 5 同时解决了这两个问题。Pop 消费模型按消息（而非按分区）管理消息生命周期，使一个队列可以同时被多个消费者共享——从而解锁横向弹性。基于 gRPC 构建的 Proxy 架构，带来了轻量级、与语言无关的客户端。但开箱即用的 RocketMQ 5 并不能覆盖美团在大规模场景下的需求：一套统一的全局元数据层，以及跨多个部署环境和业务单元的企业级流量隔离。 + +本次演讲将梳理我们在美团将 RocketMQ 5 投入生产所做出的决策与权衡。我们涵盖三个方面：第一，我们如何设计一套三层拓扑——租户（tenant）、放置组（placement group）和集群（cluster），其中放置组作为租户内的爆炸半径边界；第二，我们如何完全在 RocketMQ 之外实现多种模式的流量隔离与路由，使 broker 保持为一个纯粹的消息原语，而隔离策略可以独立演进；第三，我们如何构建单一权威的元数据服务，让所有组件保持同步。 + +我们将结合正在进行的生产迁移的真实结果来展开。大规模运行验证了我们的核心架构选型，但也暴露了一些我们仍在攻克的难题——从 RocketMQ 5 复制模型中的数据面约束，到其共享 commit log 设计中固有的容量隔离挑战。我们希望这些生产经验对更广泛的 RocketMQ 社区有所助益。 + +### 讲师: + + +
+ +Jiangge Zhang：美团软件架构师 + +Jiangge Zhang 是美团的架构师，就职于消息基础设施团队。他此前从事 SOA 中间件工作，整个职业生涯都专注于大规模分布式系统。 \ No newline at end of file diff --git a/content/sessions/messaging-1212597.md b/content/sessions/messaging-1212597.md new file mode 100644 index 000000000..6f8c9b0a3 --- /dev/null +++ b/content/sessions/messaging-1212597.md @@ -0,0 +1,37 @@ +--- +title: "Apache RocketMQ For AI: Reliable Collaboration Mechanisms for Multi-Agent Systems" +date: "2026-08-07T14:00:00" +track: "messaging" +presenters: "Zhou Li" +stype: "Chinese Session" +room: "JingYi Hall" +--- + +Section 1: Engineering Challenges in Agent Collaboration +1.1 Communication Complexity: Link fragmentation, resource contention, and scheduling blind spots under concurrent inference. +1.2 Engineering Hazards: Disrupted communication links, resource contention and scheduling failure caused by transient bursts, and the difficulty of fault recovery following the loss of collaborative context. + +Section 2: Lite-Topic: Efficient Routing and Scheduling for Massive Sessions +2.1 Dynamic Session Management: Building a TTL-based, second-level dynamic topology to achieve logical isolation of session spaces between agents. +2.2 ReadySet Scheduling Mechanism: Optimizing the delivery efficiency of massive agent events via distributed brokers, addressing "thundering herd" problems and latency jitter under ultra-high concurrency. + +Section 3: RocketMQ-A2A Asynchronous Interaction Model: Decoupling and Deterministic Control +3.1 Asynchronous Paradigm Refactoring: Building a full-link asynchronous model for "task orchestration, trigger execution, and result return," sinking collaboration logic into the RocketMQ foundation. +3.2 Deterministic Communication Orchestration: Defining standardized protocols for inter-agent communication to ensure the deterministic execution of complex business flows. + +Section 4: Collaborative Consistency and Observability: Constructing Event-Driven Evidence Chains +4.1 Semantic Enhancement and State Recovery: Leveraging idempotency design and event logs to achieve atomic interruption replay and lossless state migration for agent collaboration flows. +4.2 Full-Link Collaborative Auditing: Creating "collaborative evidence chains" based on structured messages to transform fragmented agent behaviors into traceable and attributable system governance data. + +Section 5: Large-Scale Cluster Performance and Practical Closed-Loop +5.1 Stability and fault recovery under high-pressure collaboration. +5.2 How AI gateways leverage the messaging foundation to achieve fault self-healing and collaborative task orchestration for large-scale multi-agent systems. + +### Speakers: + + +
+ +Zhou Li: aliyun + +Messaging Expert at Alibaba Cloud, responsible for Alibaba's core messaging middleware \ No newline at end of file diff --git a/content/sessions/messaging-1212597.zh.md b/content/sessions/messaging-1212597.zh.md new file mode 100644 index 000000000..16edc4645 --- /dev/null +++ b/content/sessions/messaging-1212597.zh.md @@ -0,0 +1,37 @@ +--- +title: "面向 AI 的 Apache RocketMQ：多 Agent 系统的可靠协作机制" +date: "2026-08-07T14:00:00" +track: "messaging" +presenters: "Zhou Li" +stype: "中文演讲" +room: "静宜厅" +--- + +第一部分：Agent 协作中的工程挑战 +1.1 通信复杂度：并发推理下的链路碎片化、资源竞争与调度盲区。 +1.2 工程风险：瞬时突发造成的通信链路中断、资源竞争与调度失败，以及在协作上下文丢失后故障恢复的困难。 + +第二部分：Lite-Topic：面向海量会话的高效路由与调度 +2.1 动态会话管理：构建基于 TTL 的秒级动态拓扑，实现 agent 之间会话空间的逻辑隔离。 +2.2 ReadySet 调度机制：通过分布式 broker 优化海量 agent 事件的投递效率，解决超高并发下的"惊群"（thundering herd）问题与延迟抖动。 + +第三部分：RocketMQ-A2A 异步交互模型：解耦与确定性控制 +3.1 异步范式重构：为"任务编排、触发执行、结果返回"构建全链路异步模型，将协作逻辑下沉到 RocketMQ 基础设施中。 +3.2 确定性通信编排：为 agent 之间的通信定义标准化协议，确保复杂业务流程的确定性执行。 + +第四部分：协作一致性与可观测性：构建事件驱动的证据链 +4.1 语义增强与状态恢复：利用幂等设计和事件日志，实现 agent 协作流程的原子化中断重放与无损状态迁移。 +4.2 全链路协作审计：基于结构化消息构建"协作证据链"，将碎片化的 agent 行为转化为可追溯、可归因的系统治理数据。 + +第五部分：大规模集群性能与实践闭环 +5.1 高压协作下的稳定性与故障恢复。 +5.2 AI 网关如何借助消息基础设施，实现大规模多 Agent 系统的故障自愈与协作任务编排。 + +### 讲师: + + +
+ +Zhou Li：阿里云 + +阿里云消息中间件专家，负责阿里巴巴核心消息中间件。 \ No newline at end of file diff --git a/content/sessions/messaging-1213289.md b/content/sessions/messaging-1213289.md new file mode 100644 index 000000000..1fadbbd88 --- /dev/null +++ b/content/sessions/messaging-1213289.md @@ -0,0 +1,19 @@ +--- +title: "Exploration and Practice of Low-Latency Optimizations in Apache RocketMQ" +date: "2026-08-09T14:30:00" +track: "messaging" +presenters: "Xikai Wei" +stype: "Chinese Session" +room: "JingYi Hall" +--- + +In some of our usage scenarios, strict low-latency requirements exist, with the core demand being that P99 or maximum latency is stably controlled within a fixed threshold. This talk shares the exploration and efforts our team has made in optimizing and implementing the low-latency mode of Apache RocketMQ. Starting from the core latency bottlenecks in the RocketMQ architecture, we sort out the key links affecting end-to-end latency, including network transmission, message scheduling, and resource allocation, and elaborate on the targeted optimizations, architectural adjustments, and implementation effect verification we have carried out. Additionally, this topic will share the methodologies and practical experiences of RocketMQ full-link latency tuning, hoping to provide some help to community developers and users. + +### Speakers: + + +
+ +Xikai Wei: Senior Development Engineer at Tencent Cloud + +Senior Development Engineer at Tencent Cloud, with years of experience in computing, storage, and distributed system performance optimization. \ No newline at end of file diff --git a/content/sessions/messaging-1213289.zh.md b/content/sessions/messaging-1213289.zh.md new file mode 100644 index 000000000..a7014f60a --- /dev/null +++ b/content/sessions/messaging-1213289.zh.md @@ -0,0 +1,19 @@ +--- +title: "Apache RocketMQ 低延迟优化的探索与实践" +date: "2026-08-09T14:30:00" +track: "messaging" +presenters: "熙凯 魏" +stype: "中文演讲" +room: "静宜厅" +--- + +在我们的一些使用场景中，存在严格的低延迟要求，核心诉求是将 P99 或最大延迟稳定控制在某个固定阈值之内。本次分享将介绍我们团队在 Apache RocketMQ 低延迟模式的优化与实现方面所做的探索和努力。我们从 RocketMQ 架构中的核心延迟瓶颈出发，梳理影响端到端延迟的关键环节，包括网络传输、消息调度和资源分配，并详细阐述我们针对性地进行的优化、架构调整以及实现效果的验证。此外，本议题还将分享 RocketMQ 全链路延迟调优的方法论和实践经验，希望能为社区的开发者和用户提供一些帮助。 + +### 讲师: + + +
+ +熙凯 魏: 腾讯云高级开发工程师 + +腾讯云高级开发工程师，拥有多年计算、存储及分布式系统性能优化经验。 \ No newline at end of file diff --git a/content/sessions/microservice-1196673.md b/content/sessions/microservice-1196673.md new file mode 100644 index 000000000..0a85f8be0 --- /dev/null +++ b/content/sessions/microservice-1196673.md @@ -0,0 +1,19 @@ +--- +title: "Deep Dive: Implementing OTel's Latest RPC Semantic Conventions in Apache Dubbo" +date: "2026-08-09T15:45:00" +track: "microservice" +presenters: "Zihao Rao" +stype: "Chinese Session" +room: "Mtn Yang Hall" +--- + +Over the past year, the OTel community has focused on stabilizing RPC-related semantic conventions—specifically, engaging in in-depth discussions to advance RPC framework semantics from an "experimental" to a "stable" state. This presentation outlines the latest developments in RPC semantic conventions and details the design and implementation of these semantic conventions within the Dubbo framework. + +### Speakers: + + +
+ +Zihao Rao: Passionate about Opensource + +Zihao is a software engineer at Alibaba Cloud. Over the past few years, he has participated in several well-known open source projects, he is steering committee member of Spring Cloud Alibaba project, and is an approver for OpenTelemetry Java Instrumentation now. \ No newline at end of file diff --git a/content/sessions/microservice-1196673.zh.md b/content/sessions/microservice-1196673.zh.md new file mode 100644 index 000000000..8979a4330 --- /dev/null +++ b/content/sessions/microservice-1196673.zh.md @@ -0,0 +1,19 @@ +--- +title: "深入剖析：在 Apache Dubbo 中实现 OTel 最新 RPC 语义约定" +date: "2026-08-09T15:45:00" +track: "microservice" +presenters: "Zihao Rao" +stype: "中文演讲" +room: "阳山会议室" +--- + +过去一年，OTel 社区致力于稳定 RPC 相关的语义约定——具体而言，通过深入讨论，推动 RPC 框架语义从"实验性"状态迈向"稳定"状态。本演讲将介绍 RPC 语义约定的最新进展，并详细讲解这些语义约定在 Dubbo 框架中的设计与实现。 + +### 讲师: + + +
+ +Zihao Rao：热衷于开源 + +Zihao 是阿里巴巴云的软件工程师。过去几年间，他参与了多个知名开源项目，是 Spring Cloud Alibaba 项目的指导委员会（steering committee）成员，目前也是 OpenTelemetry Java Instrumentation 的 approver。 \ No newline at end of file diff --git a/content/sessions/microservice-1206998.md b/content/sessions/microservice-1206998.md new file mode 100644 index 000000000..d71c4c871 --- /dev/null +++ b/content/sessions/microservice-1206998.md @@ -0,0 +1,27 @@ +--- +title: "Fixing Authorization in Microservices Without Breaking Everything" +date: "2026-08-09T15:15:00" +track: "microservice" +presenters: "Aram Andreasyan" +stype: "English Session" +room: "Mtn Yang Hall" +--- + +Authorization is one of those things that starts simple and quietly becomes a mess as systems grow. In a microservices setup, it often ends up duplicated across services, implemented slightly differently each time, and hard to reason about when something goes wrong. + +In this talk, Aram will walk through what typically happens as teams scale from a single service to many and how access control logic starts to drift across the system. We’ll look at common patterns that seem fine at first but break down over time, especially when services need to share context or enforce consistent rules. + +From there, he'll go over a more practical approach to handling authorization as a separate concern instead of embedding it everywhere. This includes how to structure policies, how services interact with an external decision layer, and what changes in terms of development and debugging. + +The focus is on real implementation patterns and trade-offs, not theory. By the end, the audience should have a clearer way to think about authorization in distributed systems and how to avoid the usual pitfalls without rewriting everything from scratch. + +### Speakers: + + +
+ +Aram Andreasyan: Director of Solutions + +Aram specializes in helping organizations simplify and scale authorization through policy-as-code approaches. With a strong background in cybersecurity, infrastructure, and access control, he works closely with engineering and security teams to design efficient, compliant, and developer-friendly authorization systems. +Prior to Cerbos, Aram consulted for technology startups and enterprise clients across EMEA and North America, focusing on digital transformation and cloud security strategies. Aram has experience working with distributed architectures, MDM (Master Data Management), and complex SaaS deployments. He's passionate about externalized authorization, IAM modernization, and helping economic buyers (CISOs, security architects, and engineering leaders) align technical decisions with regulatory and business needs. +When Aram's not solving authorization challenges, he enjoys staying close to tech communities, mentoring startups, and occasionally moonlight as a cyberpunk lore nerd. \ No newline at end of file diff --git a/content/sessions/microservice-1206998.zh.md b/content/sessions/microservice-1206998.zh.md new file mode 100644 index 000000000..8b5f06a3a --- /dev/null +++ b/content/sessions/microservice-1206998.zh.md @@ -0,0 +1,29 @@ +--- +title: "修复微服务中的授权问题，且不必推倒重来" +date: "2026-08-09T15:15:00" +track: "microservice" +presenters: "Aram Andreasyan" +stype: "英文演讲" +room: "阳山会议室" +--- + +授权这件事往往一开始很简单，但随着系统规模的增长，会在不知不觉中变得一团糟。在微服务架构下，授权逻辑常常在多个服务之间重复实现，每次的实现方式都略有不同，一旦出现问题，就很难理清原因。 + +在本次演讲中，Aram 将梳理团队从单一服务扩展到众多服务时通常会出现的状况，以及访问控制逻辑如何在系统中逐渐出现偏差。我们会审视那些最初看起来没问题、但随时间推移逐渐失效的常见模式，尤其是当服务之间需要共享上下文或执行一致规则时。 + +随后，他将介绍一种更实用的做法：把授权作为一个独立的关注点来处理，而不是到处嵌入到业务代码中。这包括如何组织策略（policy）、服务如何与外部的决策层（decision layer）交互，以及在开发和调试方面会发生哪些变化。 + +本次演讲聚焦于真实的实现模式与权衡，而非理论。演讲结束后，听众应当能够以更清晰的思路来思考分布式系统中的授权问题，并学会如何在不推倒重来的情况下避开常见陷阱。 + +### 讲师: + + +
+ +Aram Andreasyan：解决方案总监 + +Aram 专注于通过 policy-as-code（策略即代码）的方式，帮助组织简化并扩展其授权体系。他在网络安全、基础设施和访问控制方面拥有深厚背景，与工程团队和安全团队紧密合作，设计高效、合规且对开发者友好的授权系统。 + +在加入 Cerbos 之前，Aram 曾为 EMEA 和北美地区的技术初创公司及企业客户提供咨询服务，专注于数字化转型和云安全策略。Aram 在分布式架构、MDM（主数据管理）以及复杂的 SaaS 部署方面拥有丰富经验。他热衷于外部化授权（externalized authorization）、IAM 现代化，并帮助决策者（CISO、安全架构师和工程负责人）将技术决策与合规及业务需求对齐。 + +工作之余，Aram 喜欢与科技社区保持紧密联系、为初创公司提供指导，偶尔也会"兼职"做个赛博朋克背景故事的极客。 \ No newline at end of file diff --git a/content/sessions/microservice-1212847.md b/content/sessions/microservice-1212847.md new file mode 100644 index 000000000..53a3eecf3 --- /dev/null +++ b/content/sessions/microservice-1212847.md @@ -0,0 +1,22 @@ +--- +title: "Evolution of Open-Source Projects for Agents: Seata’s Path to AI-Friendliness" +date: "2026-08-09T14:30:00" +track: "microservice" +presenters: "Feng Zhang" +stype: "Chinese Session" +room: "Mtn Yang Hall" +--- + +In the past, open-source projects have largely built their communities and operations around human developers: clear wikis, readable code comments, and user-friendly graphical dashboards. However, with the rapid rise of large language models and AI agents, open-source codebases and infrastructure are now facing a whole new kind of user. In the future, open-source projects should not only be easy and enjoyable for humans to use, but also structured in a way that AI can “understand” and “debug” effectively. + +From the perspective of the open-source ecosystem, this analysis explores how open-source projects can become more “AI-friendly” in the age of AI. Using Apache Seata as a starting point, we will introduce our journey toward AI-oriented optimization. + +### Speakers: + + +
+ +Feng Zhang: DiDi Global Inc, Senior Software Engineer + +Open-source enthusiast +Apache Seata(incubating) Committer \ No newline at end of file diff --git a/content/sessions/microservice-1212847.zh.md b/content/sessions/microservice-1212847.zh.md new file mode 100644 index 000000000..943b03f0a --- /dev/null +++ b/content/sessions/microservice-1212847.zh.md @@ -0,0 +1,22 @@ +--- +title: "面向 Agent 的开源项目演进：Seata 走向 AI 友好之路" +date: "2026-08-09T14:30:00" +track: "microservice" +presenters: "Feng Zhang" +stype: "中文演讲" +room: "阳山会议室" +--- + +过去，开源项目主要围绕人类开发者来构建社区和运营：清晰的 wiki、易读的代码注释，以及对用户友好的图形化面板。然而，随着大语言模型和 AI agent 的快速崛起，开源代码库和基础设施正面对一种全新的"用户"。未来，开源项目不仅要让人类用起来轻松愉悦，还应当在结构上便于 AI 有效地"理解"和"调试"。 + +本演讲从开源生态的视角出发，探讨在 AI 时代开源项目如何变得更加"AI 友好"。我们将以 Apache Seata 为切入点，分享我们走向面向 AI 优化的实践历程。 + +### 讲师: + + +
+ +Feng Zhang：DiDi Global Inc（滴滴），资深软件工程师 + +开源爱好者 +Apache Seata（孵化中）Committer \ No newline at end of file diff --git a/content/sessions/microservice-1213252.md b/content/sessions/microservice-1213252.md new file mode 100644 index 000000000..345950bcb --- /dev/null +++ b/content/sessions/microservice-1213252.md @@ -0,0 +1,19 @@ +--- +title: "ServerlessX: A Kernel-Space Serverless System for RDMA-based Disaggregated Architectures" +date: "2026-08-09T14:00:00" +track: "microservice" +presenters: "Mingxuan Liu" +stype: "Chinese Session" +room: "Mtn Yang Hall" +--- + +This session explores a novel Linux kernel-space Serverless framework designed to bridge the gap between physically disaggregated hardware pools (e.g. disaggregated memory) and the instant execution semantics of modern Serverless functions. By building a three-layer architecture spanning an RDMA network foundation, distributed OS kernel primitives (RDMA Fork, Map, and mmap), and application-specific scaling strategies, we demonstrate how to achieve millisecond-level elasticity across disaggregated resources. Attendees will dive into three practical cases, including rapid GPU scale-out for LLM inference, dynamic memory scale-up for recommendation systems, and decoupled I/O capacity scaling for storage engines. + +### Speakers: + + +
+ +Mingxuan Liu: Postdoctoral Fellow, University of Macau + +Mingxuan Liu is a Postdoctoral Fellow at the University of Macau. With over a decade of research experience, his work spans operating system kernels, RDMA networking, serverless computing, and LLM infrastructure. His doctoral research focused on kernel-level serverless resource elasticity for RDMA-based disaggregated architectures. He has published more than ten first-author papers in leading CCF-A/B-ranked academic venues. At the University of Macau, he continues his research on high-performance distributed systems. \ No newline at end of file diff --git a/content/sessions/microservice-1213252.zh.md b/content/sessions/microservice-1213252.zh.md new file mode 100644 index 000000000..090008791 --- /dev/null +++ b/content/sessions/microservice-1213252.zh.md @@ -0,0 +1,19 @@ +--- +title: "ServerlessX：面向基于 RDMA 的分离式架构的内核态 Serverless 系统" +date: "2026-08-09T14:00:00" +track: "microservice" +presenters: "Mingxuan Liu" +stype: "中文演讲" +room: "阳山会议室" +--- + +本演讲探讨一种新颖的 Linux 内核态 Serverless 框架，旨在弥合物理分离的硬件资源池（例如分离式内存）与现代 Serverless 函数即时执行语义之间的差距。通过构建一个跨越 RDMA 网络、分布式操作系统内核原语（RDMA Fork、Map 和 mmap）以及面向特定应用扩展策略的三层架构，我们展示了如何在分离式资源之间实现毫秒级的弹性伸缩。听众将深入了解三个实际案例，包括面向 LLM 推理的快速 GPU 扩展、面向推荐系统的动态内存扩容，以及面向存储引擎的解耦式 I/O 容量扩展。 + +### 讲师: + + +
+ +Mingxuan Liu 刘明轩：澳门大学博士后研究员 + +Mingxuan Liu 刘明轩，现为澳门大学博士后研究员，拥有十余年的研究经验，研究方向涵盖操作系统内核、RDMA 网络、Serverless 计算和大语言模型基础设施。他在博士期间的研究聚焦于面向 RDMA 分离式架构的内核态 Serverless 资源弹性伸缩方法。他已以第一作者身份在 CCF-A/B 类高水平学术会议和期刊发表十余篇论文。目前，他在澳门大学继续从事高性能分布式系统研究。 \ No newline at end of file diff --git a/content/sessions/observability-1195519.md b/content/sessions/observability-1195519.md new file mode 100644 index 000000000..5430b30db --- /dev/null +++ b/content/sessions/observability-1195519.md @@ -0,0 +1,19 @@ +--- +title: "Apache Cloudberry: gathering statistics for queries executed by multiple PostgreSQL instances." +date: "2026-08-08T16:45:00" +track: "observability" +presenters: "Leonid Borchuk" +stype: "English Session" +room: "Mtn BaiWang Hall" +--- + +Let's say you have Apache Cloudberry cluster that consists of several PostgreSQL instances. They all execute a single query, and you want to collect general statistics on how many resources were spent on executing this query. The built-in pg_stat_statements module is no longer suitable here, since it works within the same database. I will present a solution with a similar central idea — collecting data through query execution hooks, but instead of storing it in PostgreSQL shared memory, we send raw data to an external agent process. This process is responsible for aggregating metrics, providing a unified picture across all instances, as well as additional functions, such as the forced termination of problematic sessions, while not limited to PostgreSQL: the solution can collect data from other system components. I will talk about architecture, product tasks, and share links to the repositories (Apache 2.0 License) with the code. + +### Speakers: + + +
+ +Leonid Borchuk: Developer at Yandex Cloud, Moscow. Apache Cloudberry commiter + +Team lead of the MPP postgres development team at Yandex.Cloud. He started his career with the administration of large commercial universal and analytical databases. But now he's finally writing his own. Apache Cloudberry commiter. \ No newline at end of file diff --git a/content/sessions/observability-1195519.zh.md b/content/sessions/observability-1195519.zh.md new file mode 100644 index 000000000..5b324e41d --- /dev/null +++ b/content/sessions/observability-1195519.zh.md @@ -0,0 +1,19 @@ +--- +title: "Apache Cloudberry：为多个 PostgreSQL 实例执行的查询收集统计信息" +date: "2026-08-08T16:45:00" +track: "observability" +presenters: "Leonid Borchuk" +stype: "英文演讲" +room: "百望山会议室" +--- + +假设你拥有一个由多个 PostgreSQL 实例组成的 Apache Cloudberry 集群。它们共同执行同一条查询，而你希望收集执行该查询所耗费资源的一般性统计信息。内置的 pg_stat_statements 模块在这里已经不再适用，因为它只能工作在同一个数据库之内。我将介绍一个核心思路相似的方案——同样通过查询执行钩子（hook）来采集数据，但不再将数据存放在 PostgreSQL 的共享内存中，而是把原始数据发送给一个外部的 agent 进程。该进程负责对指标进行聚合，给出跨所有实例的统一视图，并提供额外的功能，例如强制终止有问题的会话；同时它并不仅限于 PostgreSQL：该方案还能从其他系统组件采集数据。我将介绍其架构、产品层面的任务，并分享相关仓库（Apache 2.0 License）的代码链接。 + +### 讲师: + + +
+ +Leonid Borchuk：Yandex Cloud（莫斯科）开发者，Apache Cloudberry committer + +Yandex Cloud MPP PostgreSQL 开发团队的团队负责人。他的职业生涯始于大型商用通用数据库和分析数据库的管理，而如今他终于开始编写自己的数据库。Apache Cloudberry committer。 \ No newline at end of file diff --git a/content/sessions/observability-1206017.md b/content/sessions/observability-1206017.md new file mode 100644 index 000000000..b1a8cc209 --- /dev/null +++ b/content/sessions/observability-1206017.md @@ -0,0 +1,26 @@ +--- +title: "Observing LLM Applications with SkyWalking 10.4: From Performance and Cost to Quality" +date: "2026-08-09T13:30:00" +track: "observability" +presenters: "YiMing Shao" +stype: "Chinese Session" +room: "Mtn Yang Hall" +--- + +As large language models (LLMs) become increasingly integrated into production systems, traditional APM solutions are no longer sufficient to answer critical questions such as model selection, token consumption, time to first token (TTFT), inference cost, and response quality. This session introduces the AI observability capabilities of Apache SkyWalking 10.4 and demonstrates how LLM interactions can be transformed into correlated, analyzable, and actionable observability data. + +Using the SkyWalking Spring AI 1.x Agent plugin as an example, the session will demonstrate how to automatically instrument Chat, Streaming, Tool Calling, RAG Retrieval, and Vector Search operations. It will also explain how these interactions are reported according to the OpenTelemetry GenAI Semantic Conventions, including model and provider information, token usage, TTFT, request and response metadata, and other GenAI telemetry. In addition, the session will briefly introduce how the Envoy AI Gateway exports GenAI metrics and access logs via OTLP, providing infrastructure-level observability for AI traffic. + +The talk will then dive into the OAP processing pipeline. You'll learn how SkyWalking ingests traces from SkyWalking native agents, OTLP, and Zipkin, identifies the corresponding providers and models, applies configurable model pricing, estimates inference costs, and generates aggregated metrics for performance, token usage, SLA, and cost across providers and models. These insights are available immediately through out-of-the-box dashboards. + +Finally, the session will present the evolving LLM as Judge evaluation capability. SkyWalking can sample GenAI spans for quality evaluation, apply configurable rubrics to generate quality records, ratings, and evaluation metrics, and correlate the results with traces, dashboards, and alerts. Together, these capabilities provide an end-to-end AI observability workflow—from telemetry collection and performance/cost analysis to automated quality evaluation. + +### Speakers: + + +
+ +YiMing Shao: Apache SkyWalking Committer + +Apache SkyWalking Committer +Primarily focused on research in the observability domain. \ No newline at end of file diff --git a/content/sessions/observability-1206017.zh.md b/content/sessions/observability-1206017.zh.md new file mode 100644 index 000000000..a89f6cff4 --- /dev/null +++ b/content/sessions/observability-1206017.zh.md @@ -0,0 +1,33 @@ +--- +title: "使用 SkyWalking 10.4 监控 LLM 应用：从性能、成本到质量的可观测性" +date: "2026-08-09T13:30:00" +track: "observability" +presenters: "YiMing Shao" +stype: "中文演讲" +room: "阳山会议室" +--- + +随着大语言模型逐步进入生产系统，传统 APM 已不足以回答模型选择、Token 消耗、首 Token 延迟、调用成本与输出质量等关键问 + 题。本次演讲将介绍 Apache SkyWalking 10.4 的 AI 可观测性能力，展示如何将 LLM 调用转化为可关联、可分析、可运营的观测数 + 据。 + + 演讲以 SkyWalking Spring AI 1.x Agent 插件为例，说明如何自动采集 Chat、Streaming、Tool Calling、RAG Retrieval、 + Vector Search 等调用，并遵循 OpenTelemetry GenAI Semantic Conventions 上报模型、Provider、Token、TTFT 及 + 请求响应等数据。同时也会简要介绍 Envoy AI Gateway 如何从网关侧通过 OTLP 上报 GenAI 指标与访问日志，为 AI 流量提供基础 + 设施层的观测入口。 + + 随后将深入 OAP 的处理流程：SkyWalking 如何接收 SkyWalking、OTLP 与 Zipkin Trace，匹配 Provider 与 Model，加载模型定价 + 配置并计算预估成本，最终生成按 Provider 和 Model 聚合的性能、Token、SLA 与成本指标，并通过开箱即用的 Dashboard 展示。 + + 最后将介绍正在演进的 LLM as Judge 评测能力：系统可对 GenAI Span 进行采样评测，依据可配置的 Rubrics 生成质量记录、等级 + 和评分指标，并与 Trace、Dashboard 和告警关联，形成从调用采集、性能与成本分析到质量评测的完整闭环。 + +### 讲师: + + +
+ +YiMing Shao：Apache SkyWalking Committer + +Apache SkyWalking Committer +主要专注于可观测性领域的研究。 \ No newline at end of file diff --git a/content/sessions/olap-1186419.md b/content/sessions/olap-1186419.md new file mode 100644 index 000000000..f65993347 --- /dev/null +++ b/content/sessions/olap-1186419.md @@ -0,0 +1,37 @@ +--- +title: "Is Parquet Dead? The Evolution of a Columnar Giant in the AI Era" +date: "2026-08-07T17:15:00" +track: "olap" +presenters: "Gang Wu, Zehua Zou" +stype: "Chinese Session" +room: "Mtn BaiWang Hall" +--- + +Born in the Hadoop era, Apache Parquet is now over a decade old. As AI workloads demand ultra-wide tables, high-dimensional vector embeddings, and complex semi-structured data, new contenders like Lance, Vortex, and Nimble have emerged, positioning themselves as "Parquet killers." Is the industry’s most ubiquitous columnar format finally reaching its twilight? + +In this session, we will go beyond the hype to dissect the current state of Parquet and its ambitious roadmap for 2026. We will explore why Parquet’s ecosystem remains its "unbeatable moat"—spanning every major query engine to AI hubs like Hugging Face—and, more importantly, how it is undergoing a fundamental technical rebirth to meet modern challenges. + +The presentation is structured into three core pillars: + +- The Challenger Landscape: We will analyze the performance gaps that modern formats like Lance (random access) and Nimble (meta-data efficiency) aim to fill, acknowledging the limitations of "classic" Parquet in AI-centric workloads. +- The "Self-Evolution" of Parquet: We will dive deep into the latest architectural breakthroughs within the Apache Parquet community: + (1) New Primitive Power: Exploring the Variant type for JSON-like flexibility with columnar performance, and Geometry types for specialized analytics. + (2) Next-Gen Encodings: A look at ALP (Adaptive Lossless floating-point) and FSST (Fast Static Symbol Table) for extreme compression and string acceleration. + (3) Architectural Modernization: Transitioning from Thrift to FlatBuffers to solve the metadata bottleneck in tables with tens of thousands of columns. +- Cross-Project Synergy: How Parquet is collaborating with Apache Iceberg to support native column-level updates, secondary indexing, and unstructured data storage. + +### Speakers: + + +
+ +Gang Wu: Software Engineer at Singdata + +Gang is a software engineer at Singdata working on the lakehouse storage. He is a PMC member of Apache Arrow, Apache ORC, Apache Parquet, and a committer of Apache Iceberg. He is also an ASF member. Prior to Singdata, he has worked at Uber on Apache Spark and Alibaba on the MaxCompute cloud data warehouse, respectively. + + +
+ +Zehua Zou: Software Engineer at Singdata + +I'm a contributor of Arrow, Parquet, Iceberg and so on. \ No newline at end of file diff --git a/content/sessions/olap-1186419.zh.md b/content/sessions/olap-1186419.zh.md new file mode 100644 index 000000000..c65bd6dfc --- /dev/null +++ b/content/sessions/olap-1186419.zh.md @@ -0,0 +1,37 @@ +--- +title: "Parquet 已死？一位列存巨头在 AI 时代的演进" +date: "2026-08-07T17:15:00" +track: "olap" +presenters: "Gang Wu, Zehua Zou" +stype: "中文演讲" +room: "百望山会议室" +--- + +诞生于 Hadoop 时代的 Apache Parquet 如今已有十余年历史。随着 AI 工作负载对超宽表、高维向量嵌入（embedding）以及复杂半结构化数据提出需求，Lance、Vortex、Nimble 等新兴选手纷纷登场，将自己定位为"Parquet 杀手"。业界最无处不在的列式格式，是否终于走到了暮年？ + +在本次演讲中，我们将拨开炒作的迷雾，剖析 Parquet 的现状及其雄心勃勃的 2026 年路线图。我们会探讨为什么 Parquet 的生态依然是它"不可撼动的护城河"——从各大主流查询引擎到 Hugging Face 这样的 AI 枢纽——更重要的是，它如何正在经历一次根本性的技术重生，以应对现代挑战。 + +本次演讲围绕三大核心支柱展开： + +- 挑战者格局：我们将分析 Lance（随机访问）和 Nimble（元数据效率）等现代格式旨在填补的性能差距，正视"经典" Parquet 在以 AI 为中心的工作负载下的局限。 +- Parquet 的"自我进化"：我们将深入剖析 Apache Parquet 社区中的最新架构突破： + (1) 新的原语能力：探索兼具类似 JSON 的灵活性与列式性能的 Variant 类型，以及面向专门分析的 Geometry 类型。 + (2) 下一代编码：介绍 ALP（自适应无损浮点编码）与 FSST（快速静态符号表），以实现极致压缩与字符串加速。 + (3) 架构现代化：从 Thrift 迁移到 FlatBuffers，以解决拥有数万列的表的元数据瓶颈。 +- 跨项目协同：Parquet 如何与 Apache Iceberg 协作，以支持原生的列级更新、二级索引以及非结构化数据存储。 + +### 讲师: + + +
+ +Gang Wu：Singdata 软件工程师 + +Gang 是 Singdata 的软件工程师，从事湖仓存储相关工作。他是 Apache Arrow、Apache ORC、Apache Parquet 的 PMC 成员，以及 Apache Iceberg 的 committer，同时也是 ASF 成员。在加入 Singdata 之前，他曾分别在 Uber 从事 Apache Spark 相关工作，以及在阿里巴巴从事 MaxCompute 云数据仓库相关工作。 + + +
+ +Zehua Zou：Singdata 软件工程师 + +我是 Arrow、Parquet、Iceberg 等项目的贡献者。 \ No newline at end of file diff --git a/content/sessions/olap-1195516.md b/content/sessions/olap-1195516.md new file mode 100644 index 000000000..dc1bf3bf4 --- /dev/null +++ b/content/sessions/olap-1195516.md @@ -0,0 +1,20 @@ +--- +title: "We tried DPDK for network but failed. Why and what we have learned from this?" +date: "2026-08-08T14:00:00" +track: "olap" +presenters: "Leonid Borchuk" +stype: "English Session" +room: "Mtn BaiWang Hall" +--- + +The CPU frequency is practically not growing, but everything else: network/ssd/ data volumes are growing exponentially. This poses new challenges for the databases: reduce CPU costs for working with the infrastructure. The bottleneck of MPP systems is the network. What happens if the network in Greenplum becomes 10 times faster? We accelerated 5 times. Then we tried DPDK, but realized that we couldn't move forward without serious core refinement. I would share what happened to the database when the network accelerated 5 times. And also why DPDK didn't start. +What conclusions have we learned and what are we planning to do next (spoiler: all new changes will be in Apache Cloudberry). + +### Speakers: + + +
+ +Leonid Borchuk: Developer at Yandex Cloud, Moscow. Apache Cloudberry commiter + +Team lead of the MPP postgres development team at Yandex.Cloud. He started his career with the administration of large commercial universal and analytical databases. But now he's finally writing his own. Apache Cloudberry commiter. \ No newline at end of file diff --git a/content/sessions/olap-1195516.zh.md b/content/sessions/olap-1195516.zh.md new file mode 100644 index 000000000..1af0856d0 --- /dev/null +++ b/content/sessions/olap-1195516.zh.md @@ -0,0 +1,20 @@ +--- +title: "我们尝试用 DPDK 加速网络却失败了：原因以及我们的经验教训" +date: "2026-08-08T14:00:00" +track: "olap" +presenters: "Leonid Borchuk" +stype: "英文演讲" +room: "百望山会议室" +--- + +CPU 频率实际上已经不再增长，但其他一切——网络、SSD、数据量——都在呈指数级增长。这给数据库带来了新的挑战：降低与基础设施打交道时的 CPU 开销。MPP 系统的瓶颈在于网络。如果 Greenplum 中的网络变快 10 倍会发生什么？我们实现了 5 倍的加速。随后我们尝试了 DPDK，却意识到如果不做深入的内核改进，就无法继续推进。我将分享当网络加速 5 倍后，数据库发生了什么，以及为什么 DPDK 没能跑起来。 +我们从中得出了哪些结论，以及接下来打算做什么（剧透：所有新变化都将出现在 Apache Cloudberry 中）。 + +### 讲师: + + +
+ +Leonid Borchuk：Yandex Cloud（莫斯科）开发者，Apache Cloudberry committer + +Yandex Cloud MPP PostgreSQL 开发团队的团队负责人。他的职业生涯始于大型商用通用数据库和分析数据库的管理，而如今他终于开始编写自己的数据库。Apache Cloudberry committer。 \ No newline at end of file diff --git a/content/sessions/olap-1210336.md b/content/sessions/olap-1210336.md new file mode 100644 index 000000000..491373121 --- /dev/null +++ b/content/sessions/olap-1210336.md @@ -0,0 +1,21 @@ +--- +title: "Deep Dive into Parallel Query Execution Solutions for Apache Cloudberry" +date: "2026-08-07T14:30:00" +track: "olap" +presenters: "Rose Duan" +stype: "Chinese Session" +room: "Mtn BaiWang Hall" +--- + +Apache Cloudberry is an open-source Massively Parallel Processing (MPP) database designed for large-scale analytics. While MPP databases excel at distributing queries across multiple nodes, traditional single-segment query execution often leaves CPU cores underutilized. This talk explores how parallel query execution addresses this challenge by enabling multiple worker processes to collaborate on a single query within a segment. + +In this session, we'll dive deep into the architecture and implementation of parallel query execution solutions for Apache Cloudberry. This talk is ideal for database engineers, data platform architects, and anyone interested in query optimization and parallel processing. Attendees will gain practical knowledge and understand the engineering decisions behind building high-performance analytical databases. + +### Speakers: + + +
+ +Rose Duan: Apache Cloudberry Database Developer + +Apache Cloudberry contributor, database kernel developer at HashData. \ No newline at end of file diff --git a/content/sessions/olap-1210336.zh.md b/content/sessions/olap-1210336.zh.md new file mode 100644 index 000000000..7fb2aad9d --- /dev/null +++ b/content/sessions/olap-1210336.zh.md @@ -0,0 +1,21 @@ +--- +title: "深入剖析 Apache Cloudberry 的并行查询执行方案" +date: "2026-08-07T14:30:00" +track: "olap" +presenters: "Rose Duan" +stype: "中文演讲" +room: "百望山会议室" +--- + +Apache Cloudberry 是一款面向大规模分析的开源大规模并行处理（MPP）数据库。虽然 MPP 数据库擅长将查询分发到多个节点上执行，但传统的单 segment 查询执行往往会导致 CPU 核心利用率不足。本次演讲将探讨并行查询执行如何应对这一挑战——通过让多个 worker 进程在一个 segment 内协作完成同一条查询。 + +在本次演讲中，我们将深入剖析 Apache Cloudberry 并行查询执行方案的架构与实现。本演讲非常适合数据库工程师、数据平台架构师，以及所有对查询优化和并行处理感兴趣的人士。听众将获得实用的知识，并理解构建高性能分析数据库背后的工程决策。 + +### 讲师: + + +
+ +Rose Duan：Apache Cloudberry 数据库开发者 + +Apache Cloudberry contributor，HashData 数据库内核开发者。 \ No newline at end of file diff --git a/content/sessions/olap-1211106.md b/content/sessions/olap-1211106.md new file mode 100644 index 000000000..f64c9b678 --- /dev/null +++ b/content/sessions/olap-1211106.md @@ -0,0 +1,22 @@ +--- +title: "Building Lakehouse on Apache Doris at Xiaomi: Unified Computing and Cloud-Native Operations" +date: "2026-08-07T14:00:00" +track: "olap" +presenters: "Congling Xia" +stype: "Chinese Session" +room: "Mtn BaiWang Hall" +--- + +In the PB-scale data era, the separation of data lakes and data warehouses often makes enterprises use multiple different systems together, which leads to high costs, data silos, inconsistent data standards, and inefficient operation and maintenance. +Apache Doris 3.x’s improvements in lakehouse capabilities and its compute-storage separation architecture provide a way to solve this problem. This session shares how Xiaomi uses these features to replace different old engines and build a unified data analysis platform. +We’ll focus on key points like: an overview of Apache Doris’ core lakehouse features, how it’s used and the challenges we faced at Xiaomi, the pain points of advertising business and how Apache Doris helped upgrade its architecture, multi-source federated query and Catalog management, unified semantic models for BI scenarios, and how we deploy Doris on Kubernetes, along with elastic scaling and observability practices. We’ll also share our future development plans. +This session is designed to give a reference to enterprises facing similar architecture integration challenges, and help attendees learn practical experience in building a unified, efficient and stable lakehouse architecture with Apache Doris. + +### Speakers: + + +
+ +Congling Xia: Senior Software Engineer at Xiaomi, Contributor of Apache Kylin, Trino and Apache Doris + +He is a Senior Software Engineer at Xiaomi’s Data Platform. He has years of experience in big data development and the operation & maintenance of various OLAP engines. His core focus lies in enterprise-grade OLAP engine engineering and service-oriented platform capability building. He is also an open source contributor to Apache Doris, Trino, Apache Kylin, with solid hands-on experience in lakehouse architecture, unified computing and cloud-native data operations. \ No newline at end of file diff --git a/content/sessions/olap-1211106.zh.md b/content/sessions/olap-1211106.zh.md new file mode 100644 index 000000000..8834eb75b --- /dev/null +++ b/content/sessions/olap-1211106.zh.md @@ -0,0 +1,22 @@ +--- +title: "在小米基于 Apache Doris 构建湖仓：统一计算与云原生运维" +date: "2026-08-07T14:00:00" +track: "olap" +presenters: "Congling Xia" +stype: "中文演讲" +room: "百望山会议室" +--- + +在 PB 级数据时代，数据湖与数据仓库的割裂，常常迫使企业同时使用多套不同的系统，从而带来高昂的成本、数据孤岛、数据标准不一致以及低效的运维。 +Apache Doris 3.x 在湖仓能力以及存算分离架构上的改进，为解决这个问题提供了一条路径。本次分享将介绍小米如何利用这些特性替换不同的老旧引擎，构建统一的数据分析平台。 +我们将聚焦以下关键点：Apache Doris 核心湖仓特性的概览、在小米的使用方式与遇到的挑战、广告业务的痛点以及 Apache Doris 如何帮助其架构升级、多源联邦查询与 Catalog 管理、面向 BI 场景的统一语义模型，以及我们如何在 Kubernetes 上部署 Doris，连同弹性伸缩与可观测性实践。我们还会分享未来的开发计划。 +本次分享旨在为面临类似架构整合挑战的企业提供参考，帮助听众学习使用 Apache Doris 构建统一、高效、稳定的湖仓架构的实战经验。 + +### 讲师: + + +
+ +Congling Xia：小米资深软件工程师，Apache Kylin、Trino 和 Apache Doris 的 contributor + +他是小米数据平台的资深软件工程师，在大数据开发以及多种 OLAP 引擎的运维方面拥有多年经验。他的核心关注点在于企业级 OLAP 引擎工程化和服务化的平台能力建设。他同时也是 Apache Doris、Trino、Apache Kylin 的开源 contributor，在湖仓架构、统一计算和云原数据运维方面拥有扎实的实战经验。 \ No newline at end of file diff --git a/content/sessions/olap-1211809.md b/content/sessions/olap-1211809.md new file mode 100644 index 000000000..64143026f --- /dev/null +++ b/content/sessions/olap-1211809.md @@ -0,0 +1,27 @@ +--- +title: "Impala 5.0: Where lakehouse tables meet lower latency and better operability" +date: "2026-08-07T16:45:00" +track: "olap" +presenters: "Quanlong Huang" +stype: "Chinese Session" +room: "Mtn BaiWang Hall" +--- + +Apache Impala is a native query engine implemented using a massively parallel processing (MPP) architecture for open data and open table formats. + +In this session, we will share updates from the Impala community over the past year, including highlights from the upcoming 5.0 release: + - Deeper Apache Iceberg integration, including progress toward Iceberg v3 capabilities such as row lineage and deletion vectors; REST catalog support; and broader improvements to metadata handling and table maintenance. + - A major milestone for the Calcite-based query planner, including assorted optimizations and correctness fixes for edge cases, with the goal of matching or outperforming the legacy planner where it matters most. + - Catalog scalability and observability improvements, including warm failover in catalogd HA, HMS incremental event processing, and enhancements to local-catalog mode. + - Execution-side improvements such as intermediate result caching, late materialization for arrays, more accurate memory estimation, and related optimizations. + - Additional work across workload management, admissiond, OpenTelemetry integration, and other operational and ecosystem-facing features. +We will also briefly outline ongoing efforts targeted at future releases, e.g., History-based Optimizer, PIVOT/UNPIVOT support, AI Query Profile Analyzer, etc. + +### Speakers: + + +
+ +Quanlong Huang: Cloudera, Senior Staff Engineer + +Quanlong Huang is a software engineer at Cloudera. He has been contributing to the Apache Impala project for the past 8+ years. He is a committer and PMC member of Apache Impala, also a committer of Apache ORC and contributor of some other open-source projects like Apache Hive, Apache Hadoop, Apache Thrift, etc. \ No newline at end of file diff --git a/content/sessions/olap-1211809.zh.md b/content/sessions/olap-1211809.zh.md new file mode 100644 index 000000000..67215b1b1 --- /dev/null +++ b/content/sessions/olap-1211809.zh.md @@ -0,0 +1,27 @@ +--- +title: "Impala 5.0：让湖仓表获得更低时延与更优可运维性" +date: "2026-08-07T16:45:00" +track: "olap" +presenters: "Quanlong Huang" +stype: "中文演讲" +room: "百望山会议室" +--- + +Apache Impala 是一款采用大规模并行处理（MPP）架构、面向开放数据和开放表格式实现的原生查询引擎。 + +在本次演讲中，我们将分享过去一年 Impala 社区的更新，包括即将发布的 5.0 版本的亮点： + - 与 Apache Iceberg 更深度的集成，包括在行级血缘（row lineage）和删除向量（deletion vectors）等 Iceberg v3 能力上的进展；REST catalog 支持；以及元数据处理和表维护方面更广泛的改进。 + - 基于 Calcite 的查询规划器（query planner）的一个重大里程碑，包括各种优化和针对边缘场景的正确性修复，目标是在最关键的场合追平或超越旧版规划器。 + - Catalog 可扩展性与可观测性的改进，包括 catalogd HA 的热故障转移（warm failover）、HMS 增量事件处理，以及对 local-catalog 模式的增强。 + - 执行侧的改进，例如中间结果缓存、数组的延迟物化（late materialization）、更精确的内存估算以及相关优化。 + - 在负载管理、admissiond、OpenTelemetry 集成以及其他运维和生态面向特性方面的额外工作。 +我们还将简要介绍面向未来版本正在推进的工作，例如基于历史的优化器（History-based Optimizer）、PIVOT/UNPIVOT 支持、AI Query Profile Analyzer 等。 + +### 讲师: + + +
+ +Quanlong Huang：Cloudera，Senior Staff Engineer + +Quanlong Huang 是 Cloudera 的软件工程师。过去 8 年多以来，他持续为 Apache Impala 项目做贡献。他是 Apache Impala 的 committer 和 PMC 成员，也是 Apache ORC 的 committer，并参与了 Apache Hive、Apache Hadoop、Apache Thrift 等其他开源项目的贡献。 \ No newline at end of file diff --git a/content/sessions/olap-1212540.md b/content/sessions/olap-1212540.md new file mode 100644 index 000000000..ad06c1b50 --- /dev/null +++ b/content/sessions/olap-1212540.md @@ -0,0 +1,29 @@ +--- +title: "Iceberg v3 in Impala: Puffin and Deletion Vectors" +date: "2026-08-07T16:15:00" +track: "olap" +presenters: "Péter Rózsa, Zoltán Borók-Nagy" +stype: "English Session" +room: "Mtn BaiWang Hall" +--- + +Iceberg v3 introduces performance enhancements through the Puffin specification and Deletion Vectors. This session details the engineering effort to implement a high-performance C++ Puffin reader/writer within Apache Impala. We will walk through the migration path from v2 to v3, highlighting the technical challenges of bringing these features to a C++ engine. By implementing the Puffin reader/writer, Impala also gains the ability to read and write Apache Datasketches statistics. +The core of the discussion focuses on the architectural shift from position deletes to v3 deletion vectors. We will compare these two approaches in-depth, sharing implementation details and the performance gains observed during the implementation process. +We will present real-world benchmarks and discuss plans for additional v3 features for Impala. This includes how Puffin-based deletion vectors improve scan performance and how Impala can profit from engine-agnostic statistics. Attendees will leave with a clear understanding of how to leverage Iceberg v3 for high-throughput, write-heavy workloads. + + +### Speakers: + + +
+ +Péter Rózsa: Cloudera, Software Engineer + +Péter Rózsa, a Software Engineer at Cloudera since 2020, started contributing to Apache Impala in 2022. He's been working on different parts of the Impala project, recently on Iceberg integrations + + +
+ +Zoltán Borók-Nagy: Cloudera, Principal Engineer + +Zoltán is a software engineer at Cloudera, working on Apache Impala. He is also a PMC member on the project. Currently he is leading the Impala/Iceberg integration efforts. Before Cloudera, Zoltán worked on C++ static analysis tools. He is interested in distributed, massively parallel systems, databases, and performance engineering. \ No newline at end of file diff --git a/content/sessions/olap-1212540.zh.md b/content/sessions/olap-1212540.zh.md new file mode 100644 index 000000000..052b4356b --- /dev/null +++ b/content/sessions/olap-1212540.zh.md @@ -0,0 +1,29 @@ +--- +title: "Impala 中的 Iceberg v3：Puffin 与 Deletion Vectors" +date: "2026-08-07T16:15:00" +track: "olap" +presenters: "Péter Rózsa, Zoltán Borók-Nagy" +stype: "英文演讲" +room: "百望山会议室" +--- + +Iceberg v3 通过 Puffin 规范和 Deletion Vectors（删除向量）带来了性能提升。本次演讲将详细介绍在 Apache Impala 中实现高性能 C++ Puffin 读写器（reader/writer）的工程工作。我们将走过从 v2 迁移到 v3 的路径，重点讲解把这些特性引入 C++ 引擎的技术挑战。通过实现 Puffin 读写器，Impala 同时也具备了读写 Apache Datasketches 统计信息的能力。 +讨论的核心聚焦于从 position deletes 到 v3 deletion vectors 的架构转变。我们将深入对比这两种方案，分享实现细节以及在实现过程中观察到的性能收益。 +我们将展示真实的基准测试结果，并讨论 Impala 后续 v3 特性的计划。这包括基于 Puffin 的 deletion vectors 如何提升扫描性能，以及 Impala 如何从与引擎无关的（engine-agnostic）统计信息中获益。听众将清晰地了解如何利用 Iceberg v3 来应对高吞吐、写密集型的工作负载。 + + +### 讲师: + + +
+ +Péter Rózsa：Cloudera，软件工程师 + +Péter Rózsa 自 2020 年起在 Cloudera 担任软件工程师，于 2022 年开始为 Apache Impala 做贡献。他参与过 Impala 项目的多个部分，最近主要专注于 Iceberg 集成相关工作。 + + +
+ +Zoltán Borók-Nagy：Cloudera，Principal Engineer + +Zoltán 是 Cloudera 的软件工程师，从事 Apache Impala 相关工作。他也是该项目的 PMC 成员。目前他负责牵头 Impala/Iceberg 的集成工作。在加入 Cloudera 之前，Zoltán 从事 C++ 静态分析工具的开发。他感兴趣的方向包括分布式系统、大规模并行系统、数据库以及性能工程。 \ No newline at end of file diff --git a/content/sessions/olap-1212858.md b/content/sessions/olap-1212858.md new file mode 100644 index 000000000..32ccf4941 --- /dev/null +++ b/content/sessions/olap-1212858.md @@ -0,0 +1,36 @@ +--- +title: "Efficient metadata caching in Apache Impala" +date: "2026-08-07T15:00:00" +track: "olap" +presenters: "Csaba Ringhofer, Noémi Pap-Takács" +stype: "English Session" +room: "Mtn BaiWang Hall" +--- + +How to cache enough info in memory to plan queries on > 1 million file tables? Our talk will describe Apache Impala's metadata caching solution that can deal with massive Iceberg and Hive tables efficiently. + +Caching table and file metadata is key to high query throughput by allowing query planning without reaching out to other services. For huge tables (e.g. >1M files) the metadata cache becomes a massive memory hog and expensive to maintain. Impala’s current fine-grained caching and invalidation is effective for Hive tables, but doesn’t fully translate to Iceberg’s architecture. The focus of the talk is a new, cleaner and more efficient caching solution for Iceberg tables. + +Attendees can learn about: +- metadata caching problem in SQL engines +- caching quirks for classic Hive and Iceberg tables +- the effect of file system: cost of caching in Hadoop vs object stores +- minimizing memory footprint in Java +- benchmarking the overhead of updating the cache + + +### Speakers: + + +
+ +Csaba Ringhofer: Software engineer at Cloudera + +Csaba Ringhofer has been working on Apache Impala since 2017 at Cloudera. He is a Member of the Apache Impala PMC. He studied at the Budapest University of Technology and Economics. + + +
+ +Noémi Pap-Takács: Apache Impala Committer + +Noémi Pap-Takács is a software engineer at Cloudera and a committer on the Apache Impala project. Her focus lies in performance optimization and the integration of Apache Iceberg into Impala. \ No newline at end of file diff --git a/content/sessions/olap-1212858.zh.md b/content/sessions/olap-1212858.zh.md new file mode 100644 index 000000000..787d87ccb --- /dev/null +++ b/content/sessions/olap-1212858.zh.md @@ -0,0 +1,36 @@ +--- +title: "Apache Impala 中的高效元数据缓存" +date: "2026-08-07T15:00:00" +track: "olap" +presenters: "Csaba Ringhofer, Noémi Pap-Takács" +stype: "英文演讲" +room: "百望山会议室" +--- + +如何在内存中缓存足够的信息，以便对超过 100 万个文件的表进行查询规划？我们的演讲将介绍 Apache Impala 的元数据缓存方案，它能够高效地应对海量的 Iceberg 和 Hive 表。 + +缓存表与文件元数据是实现高查询吞吐的关键，因为它让查询规划无需再去访问其他服务。对于巨型表（例如超过 100 万个文件），元数据缓存会变成内存大户，维护成本高昂。Impala 现有的细粒度缓存与失效机制对 Hive 表很有效，但并不能完全套用到 Iceberg 的架构上。本次演讲的重点，是面向 Iceberg 表的一套全新、更简洁也更高效的缓存方案。 + +听众可以了解到： +- SQL 引擎中的元数据缓存问题 +- 经典 Hive 表与 Iceberg 表在缓存上的差异与特性 +- 文件系统的影响：在 Hadoop 与对象存储中缓存的成本差异 +- 在 Java 中尽量减小内存占用 +- 对更新缓存的开销进行基准测试 + + +### 讲师: + + +
+ +Csaba Ringhofer：Cloudera 软件工程师 + +Csaba Ringhofer 自 2017 年起在 Cloudera 从事 Apache Impala 相关工作。他是 Apache Impala PMC 的成员。他毕业于布达佩斯技术与经济大学（Budapest University of Technology and Economics）。 + + +
+ +Noémi Pap-Takács：Apache Impala Committer + +Noémi Pap-Takács 是 Cloudera 的软件工程师，也是 Apache Impala 项目的 committer。她的专长在于性能优化以及将 Apache Iceberg 集成到 Impala 中。 \ No newline at end of file diff --git a/content/sessions/olap-1214229.md b/content/sessions/olap-1214229.md new file mode 100644 index 000000000..c99bc7f5d --- /dev/null +++ b/content/sessions/olap-1214229.md @@ -0,0 +1,27 @@ +--- +title: "From Metadata to Semantic Analytics: An Integrated Approach for Enterprise OLAP and AI-Powered Data" +date: "2026-08-07T15:45:00" +track: "olap" +presenters: "Binhui Liu" +stype: "Chinese Session" +room: "Mtn BaiWang Hall" +--- + +Enterprise OLAP engines, BI tools, and metric systems have been in place for years. But once organizations move into natural language data Q&A, cross-engine analytics, and AI agents consuming data automatically, hidden problems surface quickly. The same metric may have different definitions across systems, tables and columns may lack consistent business meaning, and governance rules often do not travel with the analysis flow. As a result, analytics becomes hard to reuse, AI-generated answers become hard to trust, and many promising AI use cases fail to reach production. + +This talk shares practical observations from real enterprise conversations and solution design work. It focuses on the most common gaps that appear when OLAP systems evolve toward AI-driven analytics: fragmented metadata, missing semantic layers, metric definitions that cannot be reused consistently, governance that stops at system boundaries, and unstable or uncontrolled data access context for AI applications. + +I will discuss an integrated approach for open data stacks. The idea is to use a unified metadata layer to organize data objects across multiple engines and catalogs, and a semantic layer to define reusable metrics and analytical meaning, then expose these capabilities consistently to BI tools, natural language data Q&A, and agent-based applications. + +The talk will cover three questions. First, why do OLAP systems face new consistency and governance challenges after AI is introduced? Second, how should metadata, semantic layers, query engines, and AI agents work together, with clear roles in discovery, definition, access control, and execution? Third, how can architects balance ecosystem openness, implementation complexity, and user experience, so analytics can not only run, but also be reusable, explainable, and governed? + +Rather than focusing on a single query engine, this session looks at the problem from an Apache data ecosystem perspective and explores how OLAP systems can add the missing semantic and governance layer needed for more trusted analytics, more natural data interaction, and more controlled AI-driven data consumption. + +### Speakers: + + +
+ +Binhui Liu: Solutions Engineer at Datastrato, focusing on metadata, data governance, and AI-ready data infrastructure + +Levent Liu is a Solutions Engineer at Datastrato, focusing on metadata, data governance, and AI-ready data infrastructure. He works with enterprise users on open data architecture adoption, especially around metadata management, lakehouse interoperability, and AI access to governed data. His recent work centers on how metadata, semantics, and governance can support more reliable AI use cases such as natural language BI, RAG, and agent-based workflows. \ No newline at end of file diff --git a/content/sessions/olap-1214229.zh.md b/content/sessions/olap-1214229.zh.md new file mode 100644 index 000000000..9ae50d366 --- /dev/null +++ b/content/sessions/olap-1214229.zh.md @@ -0,0 +1,27 @@ +--- +title: "从元数据到语义分析：面向企业 OLAP 与 AI 驱动数据的一体化方法" +date: "2026-08-07T15:45:00" +track: "olap" +presenters: "Binhui Liu" +stype: "中文演讲" +room: "百望山会议室" +--- + +企业 OLAP 引擎、BI 工具和指标体系已经存在多年。然而一旦组织迈向自然语言数据问答、跨引擎分析，以及由 AI agent 自动消费数据的场景，隐藏的问题便会迅速浮现：同一个指标在不同系统里可能定义各异，表和列往往缺乏一致的业务含义，治理规则也常常无法随分析流程流转。结果是分析成果难以复用，AI 生成的答案难以令人信任，许多有前景的 AI 用例迟迟无法走向生产。 + +本次演讲分享来自真实企业交流与方案设计中的实践观察，聚焦于 OLAP 系统向 AI 驱动分析演进时最常见的缺口：碎片化的元数据、缺失的语义层、无法一致复用的指标定义、止步于系统边界的治理，以及面向 AI 应用的不稳定或不可控的数据访问上下文。 + +我将讨论一种面向开放数据栈的一体化方法。其思路是：用统一的元数据层跨多个引擎和 catalog 组织数据对象，用语义层定义可复用的指标与分析含义，再把这些能力一致地暴露给 BI 工具、自然语言数据问答以及基于 agent 的应用。 + +本次演讲将围绕三个问题展开。第一，引入 AI 之后，OLAP 系统为何会面临新的一致性与治理挑战？第二，元数据、语义层、查询引擎和 AI agent 应当如何分工协作，在发现、定义、访问控制和执行中各司其职？第三，架构师该如何在生态开放性、实现复杂度与用户体验之间取得平衡，使分析不仅能跑起来，还能被复用、可解释、受治理？ + +本议题不局限于单一查询引擎，而是从 Apache 数据生态的视角来看待这个问题，探讨 OLAP 系统如何补齐其缺失的语义与治理层，从而支撑更可信的分析、更自然的数据交互，以及更可控的 AI 驱动数据消费。 + +### 讲师: + + +
+ +Binhui Liu：Datastrato 解决方案工程师，专注于元数据、数据治理以及面向 AI 的数据基础设施 + +Levent Liu 是 Datastrato 的解决方案工程师，专注于元数据、数据治理以及面向 AI 的数据基础设施。他与企业用户合作推进开放数据架构的落地，尤其围绕元数据管理、湖仓互操作性，以及 AI 对受治理数据的访问。他近期的工作聚焦于元数据、语义和治理如何支撑更可靠的 AI 用例，例如自然语言 BI、RAG 和基于 agent 的工作流。 \ No newline at end of file diff --git a/content/sessions/rust-1209378.md b/content/sessions/rust-1209378.md new file mode 100644 index 000000000..d805691f1 --- /dev/null +++ b/content/sessions/rust-1209378.md @@ -0,0 +1,21 @@ +--- +title: "Full-Stack Rust × AI Native: How We Evolve Product and R&D at NonConvex" +date: "2026-08-08T16:15:00" +track: "rust" +presenters: "Qiao dan" +stype: "Chinese Session" +room: "JingYi Hall" +--- + +1. How NonConvex, as a full-stack Rust company, made its way into AI Native. +2. How we and our agents work together in Rust projects today. +3. The open-source Rust stack behind FtAi, our next-generation FinTech product. + +### Speakers: + + +
+ +Qiao dan: Director of Non-convex AI Lab；Deeply explore the integrated application of large model technology in the financial field. + +Director of Non-convex AI Lab, responsible for the development and iteration of large-scale distributed machine learning strategies in search, recommendation, and other domains. Graduated from Peking \ No newline at end of file diff --git a/content/sessions/rust-1209378.zh.md b/content/sessions/rust-1209378.zh.md new file mode 100644 index 000000000..522291cb8 --- /dev/null +++ b/content/sessions/rust-1209378.zh.md @@ -0,0 +1,21 @@ +--- +title: "全栈 Rust × AI Native：我们在 NonConvex 如何演进产品与研发" +date: "2026-08-08T16:15:00" +track: "rust" +presenters: "Qiao dan" +stype: "中文演讲" +room: "静宜厅" +--- + +1. NonConvex 作为一家全栈 Rust 公司，是如何迈向 AI Native 的。 +2. 如今我们和我们的 agent 如何在 Rust 项目中协同工作。 +3. 我们下一代金融科技（FinTech）产品 FtAi 背后的开源 Rust 技术栈。 + +### 讲师: + + +
+ +Qiao dan：Non-convex AI Lab 负责人；深入探索大模型技术在金融领域的融合应用 + +Non-convex AI Lab 负责人，负责搜索、推荐等领域大规模分布式机器学习策略的开发与迭代。毕业于北京大学。 \ No newline at end of file diff --git a/content/sessions/rust-1211660.md b/content/sessions/rust-1211660.md new file mode 100644 index 000000000..3905422c7 --- /dev/null +++ b/content/sessions/rust-1211660.md @@ -0,0 +1,23 @@ +--- +title: "Exploring Confidential Computing for Securing Web3 and AI Agent Systems" +date: "2026-08-08T16:45:00" +track: "rust" +presenters: "Zhaofeng Chen" +stype: "Chinese Session" +room: "JingYi Hall" +--- + +Apache Teaclave is a suite of Rust SDKs that provides a memory-safe framework for developing Trusted Applications in Rust, specialized for confidential computing environments such as Arm Arm TrustZone. + +Beyond offering developer APIs, it also demonstrates practical use cases in Web3 and AI agent scenarios by enabling secure key management and signing vaults, while serving as a trust anchor for device and agent authenticity. + +In this talk, we will explore how Apache Teaclave can be applied in decentralized infrastructure, autonomous agent, and edge computing scenarios. We will also examine how attestation can be used as a proof mechanism for verifying integrity, establishing trust, and enabling secure interactions across distributed systems. + +### Speakers: + + +
+ +Zhaofeng Chen: Apache Teaclave PMC Chair + +Zhaofeng Chen is the PMC Chair of Apache Teaclave, specializing in confidential computing, Web3 infrastructure security, system security, and mobile security. \ No newline at end of file diff --git a/content/sessions/rust-1211660.zh.md b/content/sessions/rust-1211660.zh.md new file mode 100644 index 000000000..80614e602 --- /dev/null +++ b/content/sessions/rust-1211660.zh.md @@ -0,0 +1,23 @@ +--- +title: "探索用于守护 Web3 与 AI Agent 系统的机密计算" +date: "2026-08-08T16:45:00" +track: "rust" +presenters: "Zhaofeng Chen" +stype: "中文演讲" +room: "静宜厅" +--- + +Apache Teaclave 是一套 Rust SDK，为用 Rust 开发可信应用（Trusted Applications）提供内存安全的框架，专门面向诸如 Arm TrustZone 这样的机密计算环境。 + +除了提供开发者 API 之外，它通过实现安全的密钥管理与签名保险库（signing vault），展示了在 Web3 和 AI agent 场景下的实际用例，同时充当设备和 agent 真实性的信任锚（trust anchor）。 + +在本次演讲中，我们将探讨 Apache Teaclave 如何应用于去中心化基础设施、自主 agent（autonomous agent）以及边缘计算场景。我们还会研究如何把远程证明（attestation）作为一种证明机制，用于校验完整性、建立信任，并在分布式系统之间实现安全交互。 + +### 讲师: + + +
+ +Zhaofeng Chen：Apache Teaclave PMC Chair + +Zhaofeng Chen 是 Apache Teaclave 的 PMC Chair，专注于机密计算（confidential computing）、Web3 基础设施安全、系统安全以及移动安全。 \ No newline at end of file diff --git a/content/sessions/rust-1213228.md b/content/sessions/rust-1213228.md new file mode 100644 index 000000000..d3f33f6b5 --- /dev/null +++ b/content/sessions/rust-1213228.md @@ -0,0 +1,21 @@ +--- +title: "Take back your control of coroutine resources" +date: "2026-08-08T15:45:00" +track: "rust" +presenters: "Ruihang Xia" +stype: "Chinese Session" +room: "JingYi Hall" +--- + +We always care about how much workload our software can handle, especially in large-scale scenarios. However the reality is not always ideal, but almost always comes with skewed workloads and fluctuating resources, challenging our software's stability and tail latency. In this session, I'll share our experience on tuning a product-ready Rust system to achieve predictable scheduling and promising self-protection in complex environments. Including failed tries, what existing Rust infrastructure can't achieve, and how we get there. + +### Speakers: + + +
+ +Ruihang Xia: Software Architect at Greptime Inc. + +Software Architect at Greptime +Apache DataFusion PMC, Apache Arrow Committer, Apache HoraeDB PPMC +https://github.com/waynexia \ No newline at end of file diff --git a/content/sessions/rust-1213228.zh.md b/content/sessions/rust-1213228.zh.md new file mode 100644 index 000000000..20ca2af90 --- /dev/null +++ b/content/sessions/rust-1213228.zh.md @@ -0,0 +1,21 @@ +--- +title: "重新掌控你的协程资源" +date: "2026-08-08T15:45:00" +track: "rust" +presenters: "Ruihang Xia" +stype: "中文演讲" +room: "静宜厅" +--- + +我们一直很关心自己的软件能扛住多大的负载，尤其是在大规模场景下。然而现实往往并不理想——几乎总是伴随着倾斜的负载和波动的资源，不断挑战着软件的稳定性与尾延迟。在本次演讲中，我将分享我们在调优一个产品级 Rust 系统时的经验，使其在复杂环境中实现可预测的调度和可靠的自保护。内容包括失败的尝试、现有 Rust 基础设施做不到的事情，以及我们是如何达成目标的。 + +### 讲师: + + +
+ +Ruihang Xia：Greptime Inc. 软件架构师 + +Greptime 软件架构师 +Apache DataFusion PMC，Apache Arrow Committer，Apache HoraeDB PPMC +https://github.com/waynexia \ No newline at end of file diff --git a/content/sessions/rust-1213258.md b/content/sessions/rust-1213258.md new file mode 100644 index 000000000..144a42688 --- /dev/null +++ b/content/sessions/rust-1213258.md @@ -0,0 +1,19 @@ +--- +title: "DORA: Let WALL-E Return to Earth" +date: "2026-08-08T14:00:00" +track: "rust" +presenters: "Leon Li" +stype: "Chinese Session" +room: "JingYi Hall" +--- + +DORA: An embodied intelligent robot development system built using the Rust programming language. This section describes DORA's architecture, operating mechanisms, and performance advantages, highlighting its many features. + +### Speakers: + + +
+ +Leon Li: DORA Community Technical Lead + +The DORA community's technical lead is responsible for the entire DORA project's R&D process and ecosystem construction and implementation. He has extensive practical experience in robot application R&D and end-to-end solution delivery, and currently focuses on the requirements breakdown, architecture design, and technology selection of embodied intelligent robot systems. \ No newline at end of file diff --git a/content/sessions/rust-1213258.zh.md b/content/sessions/rust-1213258.zh.md new file mode 100644 index 000000000..2406d0e2f --- /dev/null +++ b/content/sessions/rust-1213258.zh.md @@ -0,0 +1,19 @@ +--- +title: "DORA：让 WALL-E 回到地球" +date: "2026-08-08T14:00:00" +track: "rust" +presenters: "Leon Li" +stype: "中文演讲" +room: "静宜厅" +--- + +DORA：一套使用 Rust 编程语言构建的具身智能机器人开发系统。本部分将介绍 DORA 的架构、运行机制和性能优势，并重点展示它的诸多特性。 + +### 讲师: + + +
+ +Leon Li：DORA 社区技术负责人 + +作为 DORA 社区的技术负责人，他负责整个 DORA 项目的研发流程以及生态的建设与落地。他在机器人应用研发和端到端方案交付方面拥有丰富的实战经验，目前专注于具身智能机器人系统的需求拆解、架构设计与技术选型。 \ No newline at end of file diff --git a/content/sessions/rust-1213284.md b/content/sessions/rust-1213284.md new file mode 100644 index 000000000..a492efa56 --- /dev/null +++ b/content/sessions/rust-1213284.md @@ -0,0 +1,21 @@ +--- +title: "RustFS Scaling Laws: Performance Analysis and Insights from a more 200-Node Distributed Cluster" +date: "2026-08-08T14:30:00" +track: "rust" +presenters: "Jinghe Ma" +stype: "Chinese Session" +room: "JingYi Hall" +--- + +As a distributed object storage solution, RustFS relies on horizontal scalability to drive capacity growth. To guarantee a robust and dependable scaling mechanism, our team executed large-scale testing involving hundreds of nodes. This presentation deep dives into the performance benchmarking and analytical methodologies used during our 200+ node RustFS cluster expansion. + +### Speakers: + + +
+ +Jinghe Ma: RustFS, Core contributior + +1. The core contributor for RustFS, focusing on Docker/Helm chart development as well as issue address. +2. Was a DevOps/DevSecOps engineer. Focusing on DevOps/DevSecOps/ChatOps/open-source software supply security. Shared best practices on SECon, TiD, DevOps community summit, etc. +3. The member of the Linux Foundation Evangelist and the member of the China Cloud Native community and China DevOps community. \ No newline at end of file diff --git a/content/sessions/rust-1213284.zh.md b/content/sessions/rust-1213284.zh.md new file mode 100644 index 000000000..af9f37f17 --- /dev/null +++ b/content/sessions/rust-1213284.zh.md @@ -0,0 +1,21 @@ +--- +title: "RustFS 扩展定律：来自 200+ 节点分布式集群的性能分析与洞察" +date: "2026-08-08T14:30:00" +track: "rust" +presenters: "Jinghe Ma" +stype: "中文演讲" +room: "静宜厅" +--- + +作为一款分布式对象存储方案，RustFS 依靠横向可扩展性来驱动容量增长。为了保障一个稳健、可靠的扩展机制，我们团队执行了涉及数百个节点的大规模测试。本次演讲将深入剖析我们在 RustFS 集群扩展到 200+ 节点过程中所采用的性能基准测试与分析方法论。 + +### 讲师: + + +
+ +Jinghe Ma：RustFS，核心贡献者 + +1. RustFS 的核心贡献者，专注于 Docker/Helm chart 开发以及问题处理。 +2. 曾担任 DevOps/DevSecOps 工程师，专注于 DevOps/DevSecOps/ChatOps 以及开源软件供应链安全。曾在 SECon、TiD、DevOps 社区峰会等场合分享最佳实践。 +3. Linux Foundation Evangelist 成员，中国云原生社区与中国 DevOps 社区成员。 \ No newline at end of file diff --git a/content/sessions/rust-1240155.md b/content/sessions/rust-1240155.md new file mode 100644 index 000000000..e41228099 --- /dev/null +++ b/content/sessions/rust-1240155.md @@ -0,0 +1,21 @@ +--- +title: "Rust and AI Coding" +date: "2026-08-08T15:00:00" +track: "rust" +presenters: "Handong Zhang" +stype: "Chinese Session" +room: "JingYi Hall" +--- + +1. Why Rust is the best language for AI coding. +2. My AI coding and writing workflow. +3. Exploring the contradiction and future between AI coding and traditional craftsmanship currently faced by the open-source community. + +### Speakers: + + +
+ +Handong Zhang: Independent Consultant, Technical Writer + +Senior Independent Rust Consultant, Former Huawei Advisor, Rust Evangelist in the Chinese Community, Author of The Way of Rust Programming,（《Rust编程之道》）and Co-organizer of RustChinaConf. \ No newline at end of file diff --git a/content/sessions/rust-1240155.zh.md b/content/sessions/rust-1240155.zh.md new file mode 100644 index 000000000..8e79948d3 --- /dev/null +++ b/content/sessions/rust-1240155.zh.md @@ -0,0 +1,21 @@ +--- +title: "Rust 语言 与 AI Coding" +date: "2026-08-08T15:00:00" +track: "rust" +presenters: "汉东 张" +stype: "中文演讲" +room: "静宜厅" +--- + +1. 为什么说 Rust 语言是 AI Coding 最佳语言 +2. 我的 AI Coding 与写作工作流 +3. 探讨开源社区目前面临的 AI Coding 与 传统手艺的矛盾与未来 + +### 讲师: + + +
+ +汉东 张：独立顾问，技术作者 + +资深独立 Rust 顾问，前华为顾问，中国 Rust 社区布道者，《Rust 编程之道》作者，RustChinaConf 联合发起人。 \ No newline at end of file diff --git a/content/sessions/streaming-1191917.md b/content/sessions/streaming-1191917.md new file mode 100644 index 000000000..39399e1a6 --- /dev/null +++ b/content/sessions/streaming-1191917.md @@ -0,0 +1,20 @@ +--- +title: "Deep Dive to Flink CDC Architecture and Large-Scale Production Practices" +date: "2026-08-09T13:30:00" +track: "streaming" +presenters: "Yanquan Lv" +stype: "Chinese Session" +room: "YuanMing Hall" +--- + +Syncing hundreds of databases and tens of thousands of tables, tracking schema evolution,connecting heterogeneous systems—these are common pain points in enterprise real-time data integration. Flink CDC introduces a new Pipeline architecture that enables declarative multi-table synchronization, automatic schema evolution tracking, and one-stop data transformation. Beyond the Pipeline architecture, Flink CDC 3.6 adds support for Flink 2.2 and delivers seamless integration with mainstream databases (MySQL, PostgreSQL, Oracle) and lakehouse systems (Iceberg, Paimon, Hudi). This session will dive into the Pipeline architecture design and share performance tuning and production deployment experiences from actual production environment large-scale real-time data lake ingestion practices, helping attendees build reliable, maintainable enterprise-grade real-time data pipelines. + +### Speakers: + + +
+ +Yanquan Lv: Apache Flink Committer + +I am an Apache Flink Committer and the primary maintainer of the Flink CDC project. +Currently working on the Open Source Big Data team at Alibaba Cloud, I focus on real-time data synchronization and stream-batch unified processing. I have helped numerous enterprises implement large-scale real-time data lake ingestion solutions, with extensive experience in distributed systems and real-time computing. \ No newline at end of file diff --git a/content/sessions/streaming-1191917.zh.md b/content/sessions/streaming-1191917.zh.md new file mode 100644 index 000000000..816b03688 --- /dev/null +++ b/content/sessions/streaming-1191917.zh.md @@ -0,0 +1,20 @@ +--- +title: "深入剖析 Flink CDC 架构与大规模生产实践" +date: "2026-08-09T13:30:00" +track: "streaming" +presenters: "Yanquan Lv" +stype: "中文演讲" +room: "圆明厅" +--- + +同步数百个数据库和数万张表、追踪 schema 变更、连接异构系统——这些都是企业实时数据集成中常见的痛点。Flink CDC 引入了一套全新的 Pipeline 架构，能够实现声明式的多表同步、自动 schema 变更追踪，以及一站式的数据转换。在 Pipeline 架构之外，Flink CDC 3.6 还新增了对 Flink 2.2 的支持，并与主流数据库（MySQL、PostgreSQL、Oracle）以及湖仓系统（Iceberg、Paimon、Hudi）实现了无缝集成。本次演讲将深入讲解 Pipeline 架构设计，并分享来自真实生产环境中大规模实时数据湖入湖实践的性能调优与生产部署经验，帮助听众构建可靠、可维护的企业级实时数据管道。 + +### 讲师: + + +
+ +Yanquan Lv：Apache Flink Committer + +我是 Apache Flink Committer，也是 Flink CDC 项目的主要维护者。 +我目前在阿里云开源大数据团队工作，专注于实时数据同步与流批一体处理。我曾帮助众多企业落地大规模实时数据湖入湖方案，在分布式系统与实时计算方面经验丰富。 \ No newline at end of file diff --git a/content/sessions/streaming-1199628.md b/content/sessions/streaming-1199628.md new file mode 100644 index 000000000..38edd7561 --- /dev/null +++ b/content/sessions/streaming-1199628.md @@ -0,0 +1,27 @@ +--- +title: "Unifying Message Streaming and Data Lake: How We Built a Streaming Lakehouse Engine on Apache Pulsar" +date: "2026-08-09T14:00:00" +track: "streaming" +presenters: "Dawei Zhang" +stype: "Chinese Session" +room: "YuanMing Hall" +--- + +As organizations increasingly demand real-time analytics on streaming data, the gap between message streaming systems and data lakes becomes a critical bottleneck. BiFang is a streaming lakehouse engine built on top of Apache Pulsar that bridges this gap by deeply integrating Apache Iceberg, Apache Arrow (Flight RPC), and RocksDB KV storage. +In this talk, we will share how we extended Apache Pulsar's Broker architecture — without forking the core — to achieve unified streaming and lakehouse storage. Key topics include: +Delta Manifest Pipeline: How we generate and commit Iceberg Delta Manifests in real-time as messages flow through Pulsar, enabling sub-second data visibility in the data lake. +Arrow Flight Data Channel: How we built a high-performance columnar data read path using Arrow Flight gRPC, supporting zero-copy server-side column pruning that reduces network bandwidth by 85%+. +Two-Phase Tiered Storage: How we offload Pulsar Ledger data from AVRO_BLOCK to Parquet format using a Rust JNI Parquet Writer, achieving seamless integration with Spark, Flink, and Trino. +KV Storage with RocksDB: How we implemented primary key deduplication and CDC Changelog generation on the Broker side, enabling upsert semantics on streaming tables. + +We will also share lessons learned from running BiFang in production at scale, including threading model design, and protocol extension strategies for Apache Pulsar. +Attendees will learn practical patterns for building lakehouse capabilities on top of existing messaging systems, and gain insights into the future of streaming-lakehouse convergence. + +### Speakers: + + +
+ +Dawei Zhang: Tencent + +Dawei Zhang is a softerware engineer at Tencent, focusing on messaging and streaming infrastructure. He is a committer to Apache Pulsar and the creator of BiFang — a streaming lakehouse engine that unifies message streaming and data lake storage. He has extensive experience in distributed systems, real-time data processing, and open-source community development. \ No newline at end of file diff --git a/content/sessions/streaming-1199628.zh.md b/content/sessions/streaming-1199628.zh.md new file mode 100644 index 000000000..5336082cf --- /dev/null +++ b/content/sessions/streaming-1199628.zh.md @@ -0,0 +1,27 @@ +--- +title: "统一消息流与数据湖：我们如何在 Apache Pulsar 之上构建流式湖仓引擎" +date: "2026-08-09T14:00:00" +track: "streaming" +presenters: "Dawei Zhang" +stype: "中文演讲" +room: "圆明厅" +--- + +随着组织对流数据实时分析的需求日益增长，消息流系统与数据湖之间的鸿沟已成为一个关键瓶颈。BiFang 是一个构建在 Apache Pulsar 之上的流式湖仓引擎，通过深度集成 Apache Iceberg、Apache Arrow（Flight RPC）以及 RocksDB KV 存储，弥合了这一鸿沟。 +在本次演讲中，我们将分享如何在不对核心进行 fork 的前提下扩展 Apache Pulsar 的 Broker 架构，实现流与湖仓的统一存储。关键话题包括： +Delta Manifest 管道：我们如何在消息流经 Pulsar 时实时生成并提交 Iceberg Delta Manifest，使数据湖中的数据达到亚秒级可见。 +Arrow Flight 数据通道：我们如何使用 Arrow Flight gRPC 构建高性能的列式数据读取路径，支持服务端零拷贝列裁剪，从而将网络带宽消耗降低 85% 以上。 +两阶段分层存储：我们如何使用 Rust JNI Parquet Writer 将 Pulsar Ledger 数据从 AVRO_BLOCK 卸载为 Parquet 格式，实现与 Spark、Flink 和 Trino 的无缝集成。 +基于 RocksDB 的 KV 存储：我们如何在 Broker 侧实现主键去重和 CDC Changelog 生成，从而在流式表上支持 upsert 语义。 + +我们还将分享在大规模生产环境中运行 BiFang 所积累的经验教训，包括线程模型设计，以及针对 Apache Pulsar 的协议扩展策略。 +听众将学到在现有消息系统之上构建湖仓能力的实用模式，并对流与湖仓融合的未来趋势获得洞见。 + +### 讲师: + + +
+ +Dawei Zhang：腾讯 + +Dawei Zhang 是腾讯的软件工程师，专注于消息与流式基础设施。他是 Apache Pulsar 的 committer，也是 BiFang 的创建者——BiFang 是一个统一消息流与数据湖存储的流式湖仓引擎。他在分布式系统、实时数据处理以及开源社区发展方面拥有丰富经验。 \ No newline at end of file diff --git a/content/sessions/streaming-1199793.md b/content/sessions/streaming-1199793.md new file mode 100644 index 000000000..59afc2944 --- /dev/null +++ b/content/sessions/streaming-1199793.md @@ -0,0 +1,26 @@ +--- +title: "A New Consumer Offset Policy — 'From Now On, I Want Every Single Message'" +date: "2026-08-09T14:30:00" +track: "streaming" +presenters: "Jiunn-Yang Huang" +stype: "Chinese Session" +room: "YuanMing Hall" +--- + +Kafka's auto.offset.reset offers only earliest, latest, and none, yet none satisfies the most common need: "start consuming from now, but don't miss a single message." latest silently drops data during partition expansion or log truncation; earliest forces reprocessing of massive history; and by_duration still risks data loss due to dynamic timestamp computation. +This talk introduces a new offset reset strategy: by_start_time. It uses the consumer group's creation timestamp as a single, stable anchor point to uniformly handle every scenario — existing partitions, newly added partitions, and offset-out-of-range — all through one ListOffsetsRequest with no hidden branches. We will walk through the KIP's design evolution, its trade-offs, and the practical impact on end users. +Target Audience: +Anyone with hands-on Kafka consumer experience who has hit offset reset issues in production. If you've debugged data loss from partition expansion late at night, or been confused by the semantic differences among auto.offset.reset options, this talk is for you. +Key Takeaways: +(1) Understand how existing strategies lose data under partition expansion and log truncation. +(2) Grasp by_start_time's core design — one unified rule anchored to group creation timestamp covering all reset scenarios. +(3) Learn the KIP's design evolution and known limitations. + +### Speakers: + + +
+ +Jiunn-Yang Huang: Backend Engineer + +I’m a backend engineer who contributes to Apache Kafka. I love learning about distributed systems and event streaming—especially how they power real-time data platforms at scale. \ No newline at end of file diff --git a/content/sessions/streaming-1199793.zh.md b/content/sessions/streaming-1199793.zh.md new file mode 100644 index 000000000..8f6990a25 --- /dev/null +++ b/content/sessions/streaming-1199793.zh.md @@ -0,0 +1,26 @@ +--- +title: "一种新的消费者 Offset 策略——'从现在起，我想要每一条消息'" +date: "2026-08-09T14:30:00" +track: "streaming" +presenters: "Jiunn-Yang Huang" +stype: "中文演讲" +room: "圆明厅" +--- + +Kafka 的 auto.offset.reset 只提供 earliest、latest 和 none 三种选择，但没有一种能满足最常见的诉求："从现在开始消费，但一条消息都不要漏"。latest 在分区扩展或日志截断时会悄悄丢数据；earliest 会强制重新处理海量历史数据；而 by_duration 由于动态时间戳计算，仍然存在丢数据的风险。 +本次演讲介绍一种新的 offset 重置策略：by_start_time。它以消费组的创建时间戳作为唯一、稳定的锚点，统一处理所有场景——已有分区、新增分区，以及 offset 越界——全部通过一次 ListOffsetsRequest 完成，没有任何隐藏的分支。我们将梳理该 KIP 的设计演进、其中的权衡，以及对最终用户的实际影响。 +目标听众： +任何有 Kafka 消费者实战经验、并在生产环境中遇到过 offset 重置问题的人。如果你曾在深夜调试过因分区扩展导致的数据丢失，或是对 auto.offset.reset 各选项之间的语义差异感到困惑，这场演讲正适合你。 +核心收获： +(1) 理解现有策略在分区扩展与日志截断时是如何丢数据的。 +(2) 掌握 by_start_time 的核心设计——一条以消费组创建时间戳为锚点、覆盖所有重置场景的统一规则。 +(3) 了解该 KIP 的设计演进与已知局限。 + +### 讲师: + + +
+ +Jiunn-Yang Huang：后端工程师 + +我是一名后端工程师，也是 Apache Kafka 的贡献者。我热爱学习分布式系统和事件流——尤其是它们如何驱动大规模的实时数据平台。 \ No newline at end of file diff --git a/content/sessions/streaming-1200805.md b/content/sessions/streaming-1200805.md new file mode 100644 index 000000000..d6d86b801 --- /dev/null +++ b/content/sessions/streaming-1200805.md @@ -0,0 +1,40 @@ +--- +title: "How Post-4.0 Kafka Streams Reshapes Development and Operations" +date: "2026-08-09T15:15:00" +track: "streaming" +presenters: "PoAn Yang" +stype: "Chinese Session" +room: "YuanMing Hall" +--- + +As Apache Kafka enters the 4.0 era, not only has the underlying architecture undergone historic transformations, but Kafka Streams has also experienced a series of evolutions in developer experience and operational mechanisms. This session will guide the audience through key updates beyond Kafka Streams 4.0, analyzing how these features fundamentally reshape the lifecycle of stateful streaming applications. + +The presentation will center around three core dimensions: + +1. Coordination & Operations +The classic rebalance protocol has been with Kafka for a decade, and so has its central cost: the join barrier. Cooperative rebalancing (KIP-429) already removed the stop-the-world pause, but the barrier itself stayed. Every round still waits for the slowest member to check in before anyone learns their new assignment, and handing a task to another instance takes two of those rounds. +KIP-848 goes after the barrier itself, by folding coordination into the heartbeat that members are already sending. The assignment is computed on the broker. Different members can sit at different points of the same reassignment. None of them is waiting on the whole group. +Kafka Streams could not simply adopt it. The unit of assignment in a streams group is a task, not a partition. A task carries a state store, wants standby copies on other instances, and belongs to a topology. KIP-1071 builds on the same broker-driven design and extends it the concept: using the heartbeat mechanism to handle the rebalance. + +2. Resilience & Error Handling +Produce failures have always been part of running in production. The way Kafka Streams handled them has changed shape three times. In the early days your handler had two choices: drop the record or fail. KIP-572 then made retriable errors retry automatically by replaying the task, but it bypassed your handler. KIP-1065 hands that decision back to you: retriable errors now go through your handler, and you choose whether to retry, continue, or fail. When you decide to stop retrying, KIP-1034 gives the record somewhere to go — a dead-letter queue. + +3. Observability +If you wanted to know which step of a pipeline was slow, your only option was to wrap every lambda by hand. Some functions don't have lambda input for measurement logic. KIP-1112 closes the gap with a single injection point: one class, one config, and every processor in the topology comes back wrapped. We will use it to get per-operator timings from a running application, and point out the one configuration mistake that makes the whole mechanism silently do nothing. + +Target Audience: +Whether you are a data engineer striving to streamline topology code, an architect focused on system tracing, or an SRE pursuing zero-downtime scaling, attendees will gain practical architectural insights and code examples from this session to fully master the next-generation capabilities of Kafka Streams. + +Key Takeaways: +* Understand why the classic rebalance protocol behaved the way it did, what KIP-848 changed, and why Kafka Streams needed a protocol of its own on top. +* Learn how error handling in Kafka Streams arrived at its current shape, and how to build bounded retry and dead-letter behaviour with the 4.x API. +* See how to attach timing, logging or tracing to every processor in a topology without editing the topology itself. + +### Speakers: + + +
+ +PoAn Yang: ASF, Apache Kafka / YuniKorn committer + +I am an open-source software enthusiast, primarily focusing on Apache Kafka. I am among the top 20 contributors, and my work involves enhancing the AsyncKafkaConsumer and developing the next-generation group coordinator. \ No newline at end of file diff --git a/content/sessions/streaming-1200805.zh.md b/content/sessions/streaming-1200805.zh.md new file mode 100644 index 000000000..ba8d879dc --- /dev/null +++ b/content/sessions/streaming-1200805.zh.md @@ -0,0 +1,38 @@ +--- +title: "4.0 之后的 Kafka Streams 如何重塑开发与运维" +date: "2026-08-09T15:15:00" +track: "streaming" +presenters: "PoAn Yang" +stype: "中文演讲" +room: "圆明厅" +--- + +随着 Apache Kafka 步入 4.0 时代，不仅底层架构经历了历史性的变革，Kafka Streams 在开发者体验与运维机制上也经历了一系列演进。本次演讲将带领听众了解 Kafka Streams 4.0 之后的关键更新，分析这些特性如何从根本上重塑有状态流应用的生命周期。 + +本次演讲将围绕三个核心维度展开： + +1. 拓扑优化与数据驾驭 +我们将探讨如何编写更轻量、更具表达力的 DSL 拓扑。借助 KIP-1104，开发者可以直接从 Key 或 Value 中提取外键（foreign key），大幅减少不必要的数据冗余和 state store 开销。此外，我们还会分析 KIP-1271 和 KIP-1285（支持 Headers 的 State Store，Headers-Aware State Stores）。此前，为了防止 Trace ID 或租户标签等元数据在有状态节点中丢失，开发者往往不得不把这些系统信息"硬塞"进纯业务负载（Value）中，导致数据结构臃肿、代码冗余。这两个提案最终让 state store 能够原生保留 Headers，使开发者可以将业务逻辑与系统信息干净地解耦。 + +2. 弹性与端到端可观测性 +在生产环境中，网络波动和瞬时故障在所难免。我们将演示如何利用 KIP-1065 在 ProductionExceptionHandler 中引入的原生 RETRY 机制，构建具备自愈能力的处理管道。随后，我们将把 KIP-1112（自定义 Processor 包装，Custom Processor Wrapping）与支持 Headers 的 state store 结合起来，展示如何在整个拓扑中无缝注入 OpenTelemetry tracing 逻辑，在有状态节点之间实现真正的端到端（E2E）Tracing。 + +3. 无缝扩缩容与运维 +最后，我们将探讨大规模集群扩缩容的运维痛点。我们会介绍 KIP-1106（基于时长的 Offset 重置，Duration-Based Offset Reset）如何帮助开发者在 Tiered Storage 中处理无限保留（infinite-retention）的 topic 时精确控制数据重放范围，避免无意义的算力浪费。我们还会深入讲解 KIP-1071（Streams Rebalance 协议，Streams Rebalance Protocol），探究为什么此前的 Cooperative Rebalancing 在大集群中仍会出现"Sync Barrier"等待时延，以及因处理超时而引发的"Rebalance 风暴"。通过剖析 KIP-1071 中新引入的服务端驱动分配（server-side driven assignment）与独立的 Background Heartbeat 机制，我们将展示 Kafka 4.0 如何彻底消除扩缩容过程中的性能抖动，带领开发者迈入真正无缝的后台扩缩容时代。 + +目标听众： +无论你是力求精简拓扑代码的数据工程师、专注于系统追踪的架构师，还是追求零停机扩缩容的 SRE，听众都能从本次演讲中获得实用的架构洞见与代码示例，全面掌握 Kafka Streams 的下一代能力。 + +核心收获： +* 通过实际的 DSL 代码示例，了解如何使用新 API 消除冗余的转换逻辑，让流式应用回归干净、可维护的业务本质。 +* 掌握如何在不污染业务代码的前提下，优雅地集成端到端（E2E）Tracing 与原生错误重试机制，提升生产环境下的系统可观测性。 +* 深入理解服务端 rebalance 与独立心跳线程的底层机制，让团队在未来 Kafka 升级、日常部署或大规模集群扩展时，能够做出更有底气的架构决策。 + +### 讲师: + + +
+ +PoAn Yang：ASF，Apache Kafka / YuniKorn committer + +我是一名开源软件爱好者，主要专注于 Apache Kafka。我位列贡献者前 20 名，工作内容包括增强 AsyncKafkaConsumer 以及开发下一代 group coordinator。 \ No newline at end of file diff --git a/content/sessions/streaming-1207387.md b/content/sessions/streaming-1207387.md new file mode 100644 index 000000000..0237620e8 --- /dev/null +++ b/content/sessions/streaming-1207387.md @@ -0,0 +1,33 @@ +--- +title: "Kafka Streams 4.2.0 Dead Letter Queue: From Manual Workarounds to Built-in Transaction Safety" +date: "2026-08-09T15:45:00" +track: "streaming" +presenters: "Eric Chang" +stype: "Chinese Session" +room: "YuanMing Hall" +--- + +Handling bad records in Kafka Streams looks simple at first: catch the bad record, write it to another topic, and continue. In practice, it becomes more subtle once deserialization errors, custom exception handlers, and exactly-once processing are involved. + +This session is a practical walkthrough of KIP-1034: Dead Letter Queue in Kafka Streams, a Kafka improvement proposal created by Damien Gasparina and coauthored by Damien Gasparina, Loic Greffier, and Sebastien Viale. I am not an author of the KIP; this talk is an engineer’s reading and hands-on exploration of the feature through runnable examples. + +Before Kafka 4.2.0, adding a Dead Letter Queue often meant writing and maintaining custom exception handlers, a separate producer, producer lifecycle code, and manual header handling for source topic, partition, offset, exception type, message, and stack trace. Kafka Streams 4.2.0 introduces built-in Dead Letter Queue support that can be enabled with just a few configuration lines. Kafka Streams creates the default Dead Letter Queue record, adds the standard error headers automatically, and sends it through its internal producer path. + +We will compare a pre-KIP-1034 manual Dead Letter Queue implementation with the Kafka 4.2.0 approach, then look at why this matters for transaction safety. With a manual Dead Letter Queue producer, the Dead Letter Queue write sits outside the Kafka Streams transaction boundary. If that write is committed and the Streams transaction later aborts, the same bad input record may be retried and written to the Dead Letter Queue again. KIP-1034 closes this gap by letting Dead Letter Queue writes follow the same Kafka Streams internal write path as normal output records. + +Attendees will leave with a practical mental model for when KIP-1034 helps, what problems it solves, and how to migrate away from ad hoc Dead Letter Queue implementations in Kafka Streams. + +References: + +- [blog post about the content of this session](https://blog.unknowntpo.me/blog/kafka-kip-1034-dlq) +- [KIP-1034](https://cwiki.apache.org/confluence/display/KAFKA/KIP-1034%3A+Dead+letter+queue+in+Kafka+Streams) +And here's the link to the KIP: + +### Speakers: + + +
+ +Eric Chang: Apache Gravitino Committer, Apache Kafka contributor, and member of OpenSourceForYou, a vibrant Taiwan-based open-source community founded by Chia-Ping Tsai + +Eric Chang is an Apache Gravitino Committer and Apache Kafka contributor based in Taiwan. He is also a member of OpenSourceForYou, a vibrant Taiwan-based open-source community founded by Chia-Ping Tsai. Eric is interested in backend systems, Kafka, and distributed data processing, and he writes technical articles with runnable examples to explain practical engineering problems. His recent work explores Kafka Streams KIP-1034 and how built-in Dead Letter Queue support changes the way applications handle bad records under exactly-once processing. \ No newline at end of file diff --git a/content/sessions/streaming-1207387.zh.md b/content/sessions/streaming-1207387.zh.md new file mode 100644 index 000000000..812202572 --- /dev/null +++ b/content/sessions/streaming-1207387.zh.md @@ -0,0 +1,33 @@ +--- +title: "Kafka Streams 4.2.0 死信队列（DLQ）：从手工变通到内置的事务安全" +date: "2026-08-09T15:45:00" +track: "streaming" +presenters: "Eric Chang" +stype: "中文演讲" +room: "圆明厅" +--- + +在 Kafka Streams 中处理坏记录（bad record）乍看很简单：捕获坏记录，把它写到另一个 topic，然后继续。但在实践中，一旦牵涉到反序列化错误、自定义异常处理器和 exactly-once 处理，事情就变得更微妙了。 + +本次演讲是对 KIP-1034：Kafka Streams 中的死信队列（Dead Letter Queue）的实战讲解，这是一项由 Damien Gasparina 发起、并由 Damien Gasparina、Loic Greffier 和 Sebastien Viale 共同撰写的 Kafka 改进提案。我并非该 KIP 的作者；本次演讲是一名工程师对该特性的研读，以及借助可运行示例所做的实践探索。 + +在 Kafka 4.2.0 之前，增加一个死信队列往往意味着要编写并维护自定义的异常处理器、一个独立的 producer、producer 的生命周期代码，还要手工处理 source topic、分区、offset、异常类型、消息和堆栈跟踪等 header。Kafka Streams 4.2.0 引入了内置的死信队列支持，只需几行配置即可启用。Kafka Streams 会创建默认的死信队列记录，自动添加标准的错误 header，并通过其内部 producer 路径发送出去。 + +我们将对比 KIP-1034 之前的手工死信队列实现与 Kafka 4.2.0 的做法，然后看看为什么这对事务安全很重要。使用手工的 DLQ producer 时，死信队列的写入位于 Kafka Streams 的事务边界之外。如果该写入被提交，而随后 Streams 事务被中止，那么同一条坏输入记录可能会被重试，并再次被写入死信队列。KIP-1034 弥合了这一缺口，让死信队列的写入与正常输出记录一样，走 Kafka Streams 的内部写入路径。 + +听众离场时将获得一个实用的思维模型，了解 KIP-1034 在何时能帮上忙、它解决了哪些问题，以及如何从 Kafka Streams 中临时的（ad hoc）死信队列实现迁移过来。 + +参考资料： + +- [关于本次演讲内容的博客文章](https://blog.unknowntpo.me/blog/kafka-kip-1034-dlQ) +- [KIP-1034](https://cwiki.apache.org/confluence/display/KAFKA/KIP-1034%3A+Dead+letter+queue+in+Kafka+Streams) +以下是该 KIP 的链接： + +### 讲师: + + +
+ +Eric Chang：Apache Gravitino Committer、Apache Kafka 贡献者，以及 OpenSourceForYou（由 Chia-Ping Tsai 创立的、活跃的台湾开源社区）成员 + +Eric Chang 是驻台湾的 Apache Gravitino Committer 和 Apache Kafka 贡献者。他同时也是 OpenSourceForYou（由 Chia-Ping Tsai 创立的、活跃的台湾开源社区）的成员。Eric 对后端系统、Kafka 和分布式数据处理感兴趣，并通过带有可运行示例的技术文章来讲解实际的工程问题。他近期的工作聚焦于 Kafka Streams 的 KIP-1034，以及内置死信队列支持如何改变应用在 exactly-once 处理下处理坏记录的方式。 \ No newline at end of file diff --git a/content/sessions/streaming-1209049.md b/content/sessions/streaming-1209049.md new file mode 100644 index 000000000..f4acea363 --- /dev/null +++ b/content/sessions/streaming-1209049.md @@ -0,0 +1,26 @@ +--- +title: "From Data Chaos to Control: How a Global Telco Tamed Petabyte-Scale Challenges with Apache Iceberg" +date: "2026-08-08T15:00:00" +track: "streaming" +presenters: "Attila Turóczy" +stype: "English Session" +room: "YuanMing Hall" +--- + +When a leading telecommunications operator hit the scaling wall with their legacy Hive infrastructure, managing petabyte-scale customer data across billions of records became untenable. Their IDPR workloads suffered from slow queries, rising storage costs,partition explosion, and schema changes that broke downstream systems. + +This session explains why they chose Apache Iceberg and how it transformed their architecture, including the business and technical decision criteria used to select Iceberg over other open table formats + +Key Takeaways: +Why Iceberg: ACID guarantees, hidden partitioning, time travel, interoperability, and the executive case that secured buy-in +Architecture & Best practices: A high level architecture plus Cloudera-recommended patterns for Iceberg + Impala, partitioning strategies, compaction policies, metadata optimization, and query engine tuning for workloads +Measurable Impact: Faster queries, significant storage and infrastructure cost reductions, simpler operations, and improved regulatory compliance. + + +### Speakers: + +
+ +Attila Turóczy: Senior Director of Engineering at Cloudera + +Apache Hive, Impala and Big Data enthusiasm at Cloudera \ No newline at end of file diff --git a/content/sessions/streaming-1209049.zh.md b/content/sessions/streaming-1209049.zh.md new file mode 100644 index 000000000..1da66d653 --- /dev/null +++ b/content/sessions/streaming-1209049.zh.md @@ -0,0 +1,26 @@ +--- +title: "从数据混乱到可控：一家全球电信运营商如何用 Apache Iceberg 驯服 PB 级挑战" +date: "2026-08-08T15:00:00" +track: "streaming" +presenters: "Attila Turóczy" +stype: "英文演讲" +room: "圆明厅" +--- + +当一家领先的电信运营商在其老旧的 Hive 基础设施上撞到扩展瓶颈时，跨数十亿条记录管理 PB 级客户数据变得难以为继。他们的 IDPR 工作负载饱受查询缓慢、存储成本上升、分区爆炸，以及 schema 变更破坏下游系统等问题的困扰。 + +本次演讲将解释他们为何选择 Apache Iceberg，以及它如何改造了他们的架构，其中包括用于在其他开放表格式之间选择 Iceberg 的业务与技术决策标准。 + +核心收获： +为何选择 Iceberg：ACID 保证、隐藏分区（hidden partitioning）、time travel、互操作性，以及促成决策层认同的方案陈述。 +架构与最佳实践：高层架构，以及 Cloudera 推荐的 Iceberg + Impala 模式、分区策略、compaction 策略、元数据优化，以及面向工作负载的查询引擎调优。 +可量化的影响：更快的查询、显著的存储与基础设施成本下降、更简单的运维，以及更强的合规性。 + + +### 讲师: + +
+ +Attila Turóczy：Cloudera 工程高级总监 + +在 Cloudera 致力于 Apache Hive、Impala 及大数据领域的推广。 \ No newline at end of file diff --git a/content/sessions/streaming-1210252.md b/content/sessions/streaming-1210252.md new file mode 100644 index 000000000..dd4848efd --- /dev/null +++ b/content/sessions/streaming-1210252.md @@ -0,0 +1,20 @@ +--- +title: "Migration from Kafka to Fluss for Lake-Stream Integration: Practical Implementation" +date: "2026-08-08T14:30:00" +track: "streaming" +presenters: "zuo yan" +stype: "Chinese Session" +room: "YuanMing Hall" +--- + +The main benefits of replacing Kafka with Fluss are as follows: +Kafka uses local centralized storage, while Fluss stores cold data in Paimon. With Paimon’s S3 storage deployed on HDDs, significant storage cost savings are achieved without sacrificing query performance. + +### Speakers: + + +
+ +zuo yan: technology expert + +Zuo Yan, technology expert, is responsible for building the data center and log platform. Contributors for open-source projects such as Apache Flink, Apache Doris, Flink CDC, StarRocks, and Fluss. \ No newline at end of file diff --git a/content/sessions/streaming-1210252.zh.md b/content/sessions/streaming-1210252.zh.md new file mode 100644 index 000000000..38e2a5a28 --- /dev/null +++ b/content/sessions/streaming-1210252.zh.md @@ -0,0 +1,20 @@ +--- +title: "为湖流集成从 Kafka 迁移到 Fluss：落地实践" +date: "2026-08-08T14:30:00" +track: "streaming" +presenters: "zuo yan" +stype: "中文演讲" +room: "圆明厅" +--- + +用 Fluss 替换 Kafka 的主要收益如下： +Kafka 使用本地集中式存储，而 Fluss 把冷数据存放在 Paimon 中。借助部署在 HDD 上的 Paimon S3 存储，在不牺牲查询性能的前提下实现了显著的存储成本节省。 + +### 讲师: + + +
+ +zuo yan：技术专家 + +Zuo Yan，技术专家，负责数据中心与日志平台的建设。是 Apache Flink、Apache Doris、Flink CDC、StarRocks、Fluss 等开源项目的贡献者。 \ No newline at end of file diff --git a/content/sessions/streaming-1210323.md b/content/sessions/streaming-1210323.md new file mode 100644 index 000000000..0e150953d --- /dev/null +++ b/content/sessions/streaming-1210323.md @@ -0,0 +1,28 @@ +--- +title: "Apache Flink Agents Deep Dive" +date: "2026-08-08T14:00:00" +track: "streaming" +presenters: "Wenjin Xie" +stype: "Chinese Session" +room: "YuanMing Hall" +--- + +As agent-based technologies and applications continue to evolve, we believe that building event-driven streaming agent applications will emerge as a critical direction for the field. Against this backdrop, the Flink community has launched Flink Agents — a subproject and agent-building framework built on top of Apache Flink. +Since the release of its first version in September 2025, Flink Agents has made significant strides in functional iteration, fault-tolerant execution, ecosystem integration, and observability, all driven by close collaboration across the community. Building on Flink Agents 0.2.1, the community developed an intelligent Flink operations agent demo capable of automatically analyzing Flink job health, providing diagnostic recommendations, and attempting to resolve issues using available tools. +This talk will cover the following topics: +● Event-driven streaming agent use cases and the advantages Flink Agents brings to this space. +● The architecture and core features of Flink Agents. +● Latest advancements in Flink Agents, including support for Agent Skills, Mem0-based Long-Term memory, and enhanced Durable Execution. +● The roadmap for Flink Agents. +● A live demo of an intelligent Flink job operations agent built with Flink Agents. + +### Speakers: + + +
+ +Wenjin Xie: Alibaba Cloud Senior Development Engineer, Apache Flink Committer + +Senior development engineer in Alibaba Cloud. +Apache Flink Committer. +One of the core contributors to the Apache Flink Agents. \ No newline at end of file diff --git a/content/sessions/streaming-1210323.zh.md b/content/sessions/streaming-1210323.zh.md new file mode 100644 index 000000000..51888fb0d --- /dev/null +++ b/content/sessions/streaming-1210323.zh.md @@ -0,0 +1,28 @@ +--- +title: "Apache Flink Agents 深度解析" +date: "2026-08-08T14:00:00" +track: "streaming" +presenters: "Wenjin Xie" +stype: "中文演讲" +room: "圆明厅" +--- + +随着基于 agent 的技术和应用不断演进，我们相信构建事件驱动的流式 agent 应用将成为该领域的一个重要方向。在这一背景下，Flink 社区推出了 Flink Agents——一个构建在 Apache Flink 之上的子项目和 agent 构建框架。 +自 2025 年 9 月发布首个版本以来，在社区各方的紧密协作下，Flink Agents 在功能迭代、容错执行、生态集成和可观测性方面都取得了长足进步。基于 Flink Agents 0.2.1，社区开发了一个智能 Flink 运维 agent 示例，能够自动分析 Flink 作业健康状况、提供诊断建议，并尝试使用可用工具来解决问题。 +本次演讲将涵盖以下主题： +● 事件驱动流式 agent 的用例，以及 Flink Agents 为该领域带来的优势。 +● Flink Agents 的架构与核心特性。 +● Flink Agents 的最新进展，包括对 Agent Skills、基于 Mem0 的长期记忆（Long-Term memory）以及增强的持久化执行（Durable Execution）的支持。 +● Flink Agents 的路线图。 +● 一个用 Flink Agents 构建的智能 Flink 作业运维 agent 的现场演示。 + +### 讲师: + + +
+ +Wenjin Xie：阿里云高级开发工程师，Apache Flink Committer + +阿里云高级开发工程师。 +Apache Flink Committer。 +Apache Flink Agents 的核心贡献者之一。 \ No newline at end of file diff --git a/content/sessions/streaming-1212230.md b/content/sessions/streaming-1212230.md new file mode 100644 index 000000000..d839c268f --- /dev/null +++ b/content/sessions/streaming-1212230.md @@ -0,0 +1,43 @@ +--- +title: "The Evolution and Production Practices of Flink 2.x in Tencent" +date: "2026-08-09T16:15:00" +track: "streaming" +presenters: "Zihao Chen" +stype: "Chinese Session" +room: "YuanMing Hall" +--- + +1. Business Context and Architectural Evolution + a. Business Drivers: As Tencent's business demands for real-time Lakehouse solutions and data preprocessing for Large Language Models (LLMs) surged, the iteration of the legacy 1.x version faced challenges such as high labor costs and a growing disconnect from the open-source community. + b. Core Value of Flink 2.x: The enhancements and optimizations introduced in Flink 2.x—spanning compute-storage separation, materialized tables, batch processing optimizations, Lakehouse architectures, and AI workflows—align perfectly with Tencent's specific business scenarios. + +2. Iterative Optimization of Engine Capabilities + To foster closer integration with the open-source community, we have fully adapted a subset of our internal capabilities to the Flink 2.x framework and are gradually contributing them back to the community. + a. Compute Performance: + - Asynchronous I/O Optimization: Supports batch asynchronous I/O, boosting throughput through batching and asynchronous processing techniques. + - Lookup Join Optimization: Supports batch asynchronous Lookup Joins. + b. AI on Flink: Supports integration with the Triton Inference Server. + c. Enhanced WindowStagger Consistency: Resolves inconsistencies in window assignment behavior that previously arose when enabling the WindowStagger feature. + d. Improved O&M and Observability Capabilities: + - Native Data Sampling: Enables zero-intrusion data probing for production tasks, thereby reducing the cost of debugging and testing real-time applications. + - HistoryServer Optimization: Addresses archiving bottlenecks in large-scale production clusters, ensuring the reliable traceability of historical tasks. + +3. Implementation Practices Across Diverse Business Scenarios + a. Streaming-Batch Unification: Showcases the performance improvements achieved by upgrading streaming-batch unified scenarios to Flink 2.x. + b. Real-time Data Streams in the AI ​​Era: + - AI Functions: Enables the direct integration of AI model calls within Flink SQL to construct low-latency AI data processing pipelines. + - PyFlink Performance Optimization: Focusing on advertising feature engineering scenarios, we share how optimizing the Python UDF runtime can significantly boost feature computation efficiency. + c. Governance of Ultra-Large-Scale State: Highlights how Delta Join and BiFang are utilized to resolve issues associated with excessive state size in dual-stream Join operations. + +4. Future Outlook: From Stream Computing to Intelligent Stream Computing + a. Balancing Performance and Cost: We aim to continuously deepen the integration of incremental computing and materialized views to achieve maximum compute reuse, thereby striking the optimal balance between performance and operational cost. + b. Event-Driven AI Agents: Explore Flink Agents, evolving Flink from a standalone data processing engine into an event-driven streaming agent framework. + +### Speakers: + + +
+ +Zihao Chen: Tencent, Senior Software Development Engineer + +Zihao has been engaged in R&D related to the Flink kernel for many years. In recent years, he has mainly focused on driving the evolution of Flink 2.x and developing Flink's autoscaling capabilities to maintain the Flink engine's technological edge and improve the stability and resource utilization of Flink jobs. \ No newline at end of file diff --git a/content/sessions/streaming-1212230.zh.md b/content/sessions/streaming-1212230.zh.md new file mode 100644 index 000000000..1614d0f09 --- /dev/null +++ b/content/sessions/streaming-1212230.zh.md @@ -0,0 +1,43 @@ +--- +title: "Flink 2.x 在腾讯的演进与生产实践" +date: "2026-08-09T16:15:00" +track: "streaming" +presenters: "Zihao Chen" +stype: "中文演讲" +room: "圆明厅" +--- + +1. 业务背景与架构演进 + a. 业务驱动：随着腾讯对实时湖仓方案以及大语言模型（LLM）数据预处理的需求激增，老旧 1.x 版本的迭代面临着人力成本高企、与开源社区日益脱节等挑战。 + b. Flink 2.x 的核心价值：Flink 2.x 所引入的增强与优化——涵盖存算分离、物化表（materialized table）、批处理优化、湖仓架构以及 AI 工作流——与腾讯的具体业务场景高度契合。 + +2. 引擎能力的迭代优化 + 为了与开源社区更紧密地融合，我们已将一部分内部能力全面适配到 Flink 2.x 框架，并逐步回馈给社区。 + a. 计算性能： + - 异步 I/O 优化：支持批量异步 I/O，通过批处理与异步处理技术提升吞吐。 + - Lookup Join 优化：支持批量异步 Lookup Join。 + b. AI on Flink：支持与 Triton Inference Server 集成。 + c. 增强 WindowStagger 一致性：解决了此前在开启 WindowStagger 特性时出现的窗口分配行为不一致问题。 + d. 改进运维与可观测能力： + - 原生数据采样（Native Data Sampling）：支持对生产作业进行零侵入的数据探查，从而降低实时应用的调试与测试成本。 + - HistoryServer 优化：解决大规模生产集群中的归档瓶颈，确保历史作业的可靠可追溯。 + +3. 多样业务场景下的落地实践 + a. 流批一体：展示把流批一体场景升级到 Flink 2.x 后所获得的性能提升。 + b. AI 时代的实时数据流： + - AI Functions：支持在 Flink SQL 中直接集成 AI 模型调用，构建低时延的 AI 数据处理管道。 + - PyFlink 性能优化：聚焦广告特征工程场景，分享我们如何通过优化 Python UDF 运行时显著提升特征计算效率。 + c. 超大规模状态治理：重点介绍如何利用 Delta Join 和 BiFang 来解决双流 Join 中状态过大相关的问题。 + +4. 未来展望：从流式计算到智能流式计算 + a. 平衡性能与成本：我们希望持续深化增量计算与物化视图的结合，实现最大化的算力复用，从而在性能与运维成本之间取得最佳平衡。 + b. 事件驱动的 AI Agents：探索 Flink Agents，把 Flink 从一个独立的数据处理引擎演进为一个事件驱动的流式 agent 框架。 + +### 讲师: + + +
+ +Zihao Chen：腾讯，资深软件开发工程师 + +Zihao 多年来一直从事 Flink 内核相关的研发。近年来，他主要专注于推动 Flink 2.x 的演进，以及开发 Flink 的自动扩缩容（autoscaling）能力，以保持 Flink 引擎的技术领先性，并提升 Flink 作业的稳定性与资源利用率。 \ No newline at end of file diff --git a/content/sessions/streaming-1213500.md b/content/sessions/streaming-1213500.md new file mode 100644 index 000000000..cfeedcaa2 --- /dev/null +++ b/content/sessions/streaming-1213500.md @@ -0,0 +1,19 @@ +--- +title: "Fluss × Alipay: Streaming Storage Practices in Marketing Scenarios" +date: "2026-08-08T15:45:00" +track: "streaming" +presenters: "Yufeng Zhu" +stype: "Chinese Session" +room: "YuanMing Hall" +--- + +This article presents the implementation of Fluss in Alipay’s marketing scenarios. Starting from business challenges, it provides a detailed account of the entire process, including technology selection, architectural design, and the deployment of core use cases. By comparing solutions such as SLS, it demonstrates Fluss’s advantages in real-time data processing. Actual performance metrics and cost optimization results are also shared, concluding with best practices to offer valuable insights for industry peers. + +### Speakers: + + +
+ +Yufeng Zhu: Senior Development Engineer at Ant Group, Apache Paimon & Fluss Contributor + +For many years, I have been dedicated to the field of real-time big data computing and data lakes, responsible for exploring the application of Paimon and Fluss in Ant's business scenarios. \ No newline at end of file diff --git a/content/sessions/streaming-1213500.zh.md b/content/sessions/streaming-1213500.zh.md new file mode 100644 index 000000000..2eaa6125d --- /dev/null +++ b/content/sessions/streaming-1213500.zh.md @@ -0,0 +1,20 @@ +--- +title: "Fluss × Alipay：营销场景下的流式存储实践" +date: "2026-08-08T15:45:00" +track: "streaming" +presenters: "育锋 朱" +stype: "中文演讲" +room: "圆明厅" +--- + +本文介绍 Fluss 在 Alipay 营销场景中的落地实践。从业务挑战出发，详细讲述整个过程，包括技术选型、架构设计，以及核心用例的部署。通过与 SLS 等方案的对比，展示 Fluss 在实时数据处理方面的优势。文中还分享了真实的性能指标与成本优化结果，最后总结最佳实践，为业界同行提供有价值的参考。 + +### 讲师: + + +
+ + +育锋 朱: 蚂蚁集团高级开发工程师、Apache Paimon & Fluss Contributor + +多年来专注于大数据实时计算/数据湖领域，负责Paimon、Fluss在蚂蚁业务场景的探索应用。 \ No newline at end of file diff --git a/content/sessions/streaming-1213708.md b/content/sessions/streaming-1213708.md new file mode 100644 index 000000000..08518e32b --- /dev/null +++ b/content/sessions/streaming-1213708.md @@ -0,0 +1,29 @@ +--- +title: "GeaFlow | Streaming Memory: Building a Real-time Stateful Backend to Empower AI+Graph Scenarios" +date: "2026-08-08T16:15:00" +track: "streaming" +presenters: "Litao Lin, Yao Zhongqiang" +stype: "Chinese Session" +room: "YuanMing Hall" +--- + +In 2026, AI Agent memory architecture is undergoing a paradigm shift. Traditional AI Agent memory systems largely adopt a batch processing mode characterized by "append-only writes + asynchronous indexing." This results in significant latency in memory updates, making it difficult to support contextual evolution in real-time interaction scenarios. The essence of memory is not a static dataset, but an unbounded data stream composed of dialogue, perception, and decision-making. How to perform low-latency state management and incremental computation on this stream represents a new challenge for stream computing technology in the AI era. +This session will focus on how stream computing technology addresses the memory challenge. Apache GeaFlow (Incubating), an open-source streaming graph computation engine, is positioned as a real-time incremental graph computation layer, serving as the real-time stateful backend for Agents. Unlike traditional graph databases and stream computing engines that lack native relationship modeling capabilities, GeaFlow processes incoming data as continuous graph updates via incremental computation—where each event triggers only necessary subgraph changes, thereby avoiding expensive full recomputations. +Specifically, GeaFlow replaces full computation with incremental streaming computation to enable real-time updates of local subgraphs the instant a dialogue occurs. Through dynamic graph state management, it maintains temporal memory as stream state, resolving state consistency issues in long-running dialogues. Furthermore, it leverages a distributed snapshot mechanism to ensure state fault tolerance and recovery for enterprise-level multi-Agent collaboration. +In the technical practice segment, we will demonstrate how to build a knowledge assistant capable of millisecond-level memory updates. We will also showcase how GeaFlow provides foundational support for AI systems operating on continuously evolving knowledge graphs. + +### Speakers: + + +
+ +Litao Lin: Apache GeaFlow (Incubating) Committer + +Apache GeaFlow (Incubating) Committer. As a core member of the project, he participated in the architecture design and development of the GeaFlow graph computing engine from scratch, with a particular focus on the design and implementation of the graph computing DSL and the evolution of data intelligence technologies. He is deeply involved in open source community activities and currently works at Ant Group. + + +
+ +Yao Zhongqiang: Graph Computing Expert & Development Engineer at Ant Group + +Deeply involved in the big data field, specializing in graph computing, real-time computing, and OLAP. Current team focus is on Agent Memory and MARL. \ No newline at end of file diff --git a/content/sessions/streaming-1213708.zh.md b/content/sessions/streaming-1213708.zh.md new file mode 100644 index 000000000..1e8adfe58 --- /dev/null +++ b/content/sessions/streaming-1213708.zh.md @@ -0,0 +1,29 @@ +--- +title: "GeaFlow | 流式记忆（Streaming Memory）：构建实时有状态后端，赋能 AI+Graph 场景" +date: "2026-08-08T16:15:00" +track: "streaming" +presenters: "Litao Lin, Yao Zhongqiang" +stype: "中文演讲" +room: "圆明厅" +--- + +2026 年，AI Agent 的记忆架构正在经历一次范式转变。传统 AI Agent 的记忆系统大多采用"只追加写入 + 异步索引"的批处理模式。这导致记忆更新存在显著时延，难以支撑实时交互场景下的上下文演进。记忆的本质并非一个静态数据集，而是由对话、感知和决策构成的无界数据流。如何在这条流上进行低时延的状态管理与增量计算，是流计算技术在 AI 时代面临的新挑战。 +本次演讲将聚焦流计算技术如何应对记忆这一挑战。Apache GeaFlow（孵化中）是一款开源的流式图计算引擎，定位为实时增量图计算层，作为 Agent 的实时有状态后端。与缺乏原生关系建模能力的传统图数据库和流计算引擎不同，GeaFlow 通过增量计算把到达的数据作为连续的图更新来处理——每个事件只触发必要的子图变更，从而避免昂贵的全量重算。 +具体而言，GeaFlow 用增量流式计算取代全量计算，使局部子图能够在对话发生的瞬间实时更新。通过动态的图状态管理，它把时序记忆作为流状态来维护，解决长会话中的状态一致性问题。此外，它借助分布式快照机制，保障企业级多 Agent 协作下的状态容错与恢复。 +在技术实践环节，我们将演示如何构建一个具备毫秒级记忆更新能力的知识助手。我们还将展示 GeaFlow 如何为运行在持续演进知识图谱之上的 AI 系统提供基础支撑。 + +### 讲师: + + +
+ +Litao Lin：Apache GeaFlow（孵化中）Committer + +Apache GeaFlow（孵化中）Committer。作为项目的核心成员，他从零参与了 GeaFlow 图计算引擎的架构设计与开发，尤其专注于图计算 DSL 的设计与实现，以及数据智能技术的演进。他深度参与开源社区活动，目前就职于 Ant Group。 + + +
+ +Yao Zhongqiang：Ant Group 图计算专家与开发工程师 + +深耕大数据领域，专长于图计算、实时计算和 OLAP。所在团队目前专注于 Agent Memory 与 MARL。 \ No newline at end of file diff --git a/content/sessions/webserver-1213943.md b/content/sessions/webserver-1213943.md new file mode 100644 index 000000000..39e88d3c0 --- /dev/null +++ b/content/sessions/webserver-1213943.md @@ -0,0 +1,19 @@ +--- +title: "Orchestrating and Diagnosing: Seata-Go Agent Platform and seata-ctl Toolchain" +date: "2026-08-09T16:15:00" +track: "webserver" +presenters: "Tun Guo" +stype: "Chinese Session" +room: "Mtn Yang Hall" +--- + +To address the complexity of long-transaction orchestration and the challenges of distributed transaction diagnostics, this talk will cover the evolution of Seata Go and its infrastructure, the capabilities of the Agent orchestration platform and diagnostic tools, and practical use cases integrating the two. + +### Speakers: + + +
+ +Tun Guo: Backend Developer, Apache Seata(Incubator) Committer + +A backend development engineer at an embedded AI company, and an Apache Seata (Incubator) Committer. \ No newline at end of file diff --git a/content/sessions/webserver-1213943.zh.md b/content/sessions/webserver-1213943.zh.md new file mode 100644 index 000000000..a5000cec0 --- /dev/null +++ b/content/sessions/webserver-1213943.zh.md @@ -0,0 +1,19 @@ +--- +title: "编排与诊断：Seata-Go Agent 平台与 seata-ctl 工具链" +date: "2026-08-09T16:15:00" +track: "webserver" +presenters: "Tun Guo" +stype: "中文演讲" +room: "阳山会议室" +--- + +为了应对长事务编排的复杂性以及分布式事务诊断的挑战，本次演讲将介绍 Seata Go 及其基础设施的演进、Agent 编排平台与诊断工具的能力，以及将二者结合的实践用例。 + +### 讲师: + + +
+ +Tun Guo：后端开发者，Apache Seata（孵化中）Committer + +一家嵌入式 AI 公司的后端开发工程师，Apache Seata（孵化中）Committer。 \ No newline at end of file diff --git a/content/tracks/5minstalk.md b/content/tracks/5minstalk.md new file mode 100644 index 000000000..805f152e9 --- /dev/null +++ b/content/tracks/5minstalk.md @@ -0,0 +1,6 @@ +--- +title: Lightning Talk +trackchairs: Richard Lin +--- + +Get ready for an exciting lightning talk session! In this fast-paced and dynamic segment, dozens of speakers will take the stage, each delivering inspiring ideas and stories in just 5 minutes. It's a fantastic opportunity to experience diverse thoughts, perspectives, and innovations. Don’t miss this chance to get inspired and engaged through a rapid succession of impactful presentations! \ No newline at end of file diff --git a/content/tracks/5minstalk.zh.md b/content/tracks/5minstalk.zh.md new file mode 100644 index 000000000..f2459c0da --- /dev/null +++ b/content/tracks/5minstalk.zh.md @@ -0,0 +1,6 @@ +--- +title: 闪电演讲 +trackchairs: "林旅强" +--- + +准备好迎接一场激动人心的闪电演讲环节吧！在这个快节奏和充满活力的环节中，数十位演讲者将登上舞台，每人用 5 分钟时间为大家带来各种有启发的观点和故事。这是一个体验多样化思想、视角和创新的绝佳机会。不要错过这个通过快速连续的影响力演讲来获得启发和参与的机会！ \ No newline at end of file diff --git a/content/tracks/agenticcoding.md b/content/tracks/agenticcoding.md new file mode 100644 index 000000000..51fb26bb0 --- /dev/null +++ b/content/tracks/agenticcoding.md @@ -0,0 +1,8 @@ +--- +title: Agentic coding +trackchairs: "tison, Xuanwo" +--- + +Agentic coding is moving from demos into production. Across Apache projects, maintainers are asking: How can we integrate LLM-powered agents into our development workflows without compromising code quality, reviewability, or the ASF’s community-driven principles? From code generation and test writing to refactoring, debugging, code review assistance, and release automation, projects across the foundation are experimenting with agents across the entire software lifecycle. These early efforts reveal not only new possibilities but also real friction points. + +This track focuses on hands‑on experiences from ASF contributors who are building, integrating, or evaluating code agents. We want to hear what actually works when agents participate in discussions, propose patches, or help with large‑scale maintenance. What breaks, what becomes easier, and what repeatable patterns are emerging? We’re looking for talks that share practical lessons to help the community turn agents from flashy helpers into dependable collaborators. \ No newline at end of file diff --git a/content/tracks/agenticcoding.zh.md b/content/tracks/agenticcoding.zh.md new file mode 100644 index 000000000..5d535d19c --- /dev/null +++ b/content/tracks/agenticcoding.zh.md @@ -0,0 +1,8 @@ +--- +title: Agentic coding +trackchairs: "陈梓立, Xuanwo" +--- + +自主编码（Agentic coding）正在从演示走向生产。在Apache项目中，维护者们在问：我们如何在不妥协代码质量、可审查性或ASF社区驱动原则的情况下，将LLM驱动的代理集成到我们的开发工作流程中？从代码生成和测试编写到重构、调试、代码审查辅助和发布自动化，基金会中的项目正在整个软件生命周期中尝试使用代理。这些早期的努力不仅揭示了新的可能性，还暴露了真实的摩擦点。 + +这个专题关注的是ASF贡献者们的实践经验，他们正在构建、集成或评估代码代理。我们希望听到代理在参与讨论、提出补丁或帮助进行大规模维护时实际有效的经验。什么会出错，什么变得更容易，以及正在出现哪些可重复的模式？我们正在寻找分享实用经验的演讲，帮助社区将代理从炫目的助手转变为可靠的合作伙伴。 \ No newline at end of file diff --git a/content/tracks/aiinfra.md b/content/tracks/aiinfra.md new file mode 100644 index 000000000..f4ed1eb5e --- /dev/null +++ b/content/tracks/aiinfra.md @@ -0,0 +1,11 @@ +--- +title: AI Infra +trackchairs: "Jerry Tan, Junping Du, Gabriel Wang" +--- + +The AI Sub-forum is a professional exchange platform focusing on the integration of artificial intelligence (AI) technology and Apache open-source projects. This sub-forum aims to bring together developers, researchers, and industry users from around the world to explore the application and development of AI technology in the open-source ecosystem, showcase cutting-edge technologies, share practical experience, and promote the implementation of open-source AI solutions across various industries. Topics suitable for this sub-forum include: + +* AI-related projects under the Apache Foundation (e.g., focusing on open-source AI frameworks and foundational libraries such as Apache TVM, Mahout, Singa, SystemML, etc.). +* Optimization of individual Apache projects in AI scenarios (e.g., projects like Spark MLlib, Flink ML). + +* AI solutions in industrial scenarios based on combinations of multiple Apache projects (e.g., methods for building AI businesses by certain enterprises using a combination of Apache projects). diff --git a/content/tracks/aiinfra.zh.md b/content/tracks/aiinfra.zh.md new file mode 100644 index 000000000..b0d1bc04c --- /dev/null +++ b/content/tracks/aiinfra.zh.md @@ -0,0 +1,15 @@ +--- +title: 人工智能基础设施 +trackchairs: "谭中意, 堵俊平, 王臣汉" +--- + +AI分论坛是一个专注于人工智能（AI）技术与Apache开源项目结合的专业交流平台。该分论坛旨在汇聚全球开发者、研究人员和行业用户，共同探讨AI技术在开源生态中的应用与发展，展示前沿技术、分享实践经验，并推动开源AI解决方案在各行业的落地。 + +适合本论坛的议题为： + +* Apache基金会下的AI相关项目（例如聚焦于开源AI框架和基础库项目Apache TVM、Mahout、Singa、SystemML等） + +* Apache单个项目在AI场景下的优化（例如Spark MLib，Flink ML等项目） + +* 基于Apache多个项目组合后在工业场景的AI解决方案例如某某企业的AI业务构建方法 + diff --git a/content/tracks/beam.md b/content/tracks/beam.md deleted file mode 100755 index 3b4f99528..000000000 --- a/content/tracks/beam.md +++ /dev/null @@ -1,3 +0,0 @@ ---- -title: Beam Summit ---- diff --git a/content/tracks/bigdata.md b/content/tracks/bigdata.md deleted file mode 100755 index 36909761a..000000000 --- a/content/tracks/bigdata.md +++ /dev/null @@ -1,3 +0,0 @@ ---- -title: Big Data ---- diff --git a/content/tracks/cassandra.md b/content/tracks/cassandra.md deleted file mode 100755 index c189d9a24..000000000 --- a/content/tracks/cassandra.md +++ /dev/null @@ -1,3 +0,0 @@ ---- -title: Apache Cassandra ---- diff --git a/content/tracks/cloudstack.md b/content/tracks/cloudstack.md deleted file mode 100755 index a8c793a9b..000000000 --- a/content/tracks/cloudstack.md +++ /dev/null @@ -1,3 +0,0 @@ ---- -title: Cloudstack Collaboration Conference ---- diff --git a/content/tracks/community.md b/content/tracks/community.md old mode 100755 new mode 100644 index fc6c9c6a4..cf6dfd652 --- a/content/tracks/community.md +++ b/content/tracks/community.md @@ -1,3 +1,10 @@ --- title: Community +trackchairs: "Yin Xu, Richard Lin" --- + +Why ASF believes "Community over Code"? There is more to open source than just code contribution. + +The Community track covers topics such as community governance model, growing open source project communities, diversity & inclusion, measuring community health, community management tools and data, project roadmaps, case studies, and any other topic around sustaining open source and open source communities. + +We welcome you to share your community story with a broader audience. diff --git a/content/tracks/community.zh.md b/content/tracks/community.zh.md new file mode 100644 index 000000000..de846c19a --- /dev/null +++ b/content/tracks/community.zh.md @@ -0,0 +1,11 @@ +--- +title: 开源社区 +trackchairs: "许银, 林旅强" +--- + + +为什么 ASF 相信 "社区重于代码"？开源不仅仅是代码贡献。 + +社区专场涵盖的主题包括：社区治理模式、发展开源项目社区、多样性和包容性、衡量社区健康、社区管理工具与数据、项目路线图、案例研究，以及其他任何关于开源和开源社区可持续性的主题。 + +我们欢迎你与更多的人分享你的社区故事。 diff --git a/content/tracks/contentdelivery.md b/content/tracks/contentdelivery.md deleted file mode 100755 index 7778cfb58..000000000 --- a/content/tracks/contentdelivery.md +++ /dev/null @@ -1,3 +0,0 @@ ---- -title: Content Delivery ---- diff --git a/content/tracks/dataai.md b/content/tracks/dataai.md new file mode 100644 index 000000000..e9327bb8c --- /dev/null +++ b/content/tracks/dataai.md @@ -0,0 +1,14 @@ +--- +title: Data + AI +trackchairs: "Juan Pan, Sheng Wu, Jeff Feng" +--- + +In the era of Generative AI and Autonomous Agents, data is no longer just a static asset—it is the dynamic fuel for intelligence. The Data + AI track explores the profound convergence where Apache’s battle-tested data infrastructure meets cutting-edge AI capabilities. + +We focus on two critical evolutionary paths: + +- Infrastructure for AI: How Apache projects are evolving to support the massive scale required by LLMs—covering vector retrieval, real-time RAG pipelines, and unstructured data governance. + +- AI-Evolved Data Systems: How AI agents and models are redefining data engineering itself—from autonomous query optimization and self-healing pipelines to AI-native features within established projects. + +This track is designed for engineers and architects who move beyond the hype. Join us to dissect concrete architectures, production-grade agentic workflows, and the future of open standards in building reliable, governed, and scalable intelligent systems. diff --git a/content/tracks/dataai.zh.md b/content/tracks/dataai.zh.md new file mode 100644 index 000000000..505dbdfe0 --- /dev/null +++ b/content/tracks/dataai.zh.md @@ -0,0 +1,14 @@ +--- +title: Data + AI +trackchairs: "潘娟, 吴晟, 冯嘉" +--- + +在生成式 AI 与自主 Agent 爆发的时代，数据不再仅仅是静态资产，而是驱动智能的动态燃料。Data + AI 版块致力于探索 Apache 经过实战检验的数据基础设施与前沿 AI 能力的深度融合。 + +我们将聚焦两大核心演进路径： + +- 面向 AI 的新基建 (Infrastructure for AI)： 探讨 Apache 数据生态如何演进以支撑 LLM 所需的海量规模——涵盖向量检索、实时 RAG 管道以及非结构化数据治理。 + +- AI 驱动的数据系统 (AI-Evolved Data Systems)： 展示 AI Agent 与大模型如何重塑数据工程本身——从自动化的查询优化、自愈合的数据管道，到现有项目中的 AI 原生特性落地。 + +本版块专为追求技术深度的工程师与架构师打造。我们将跨越炒作，深入剖析具体的架构设计、生产级 Agent 工作流，以及在构建可靠、合规、可扩展的智能系统时，开源标准所扮演的关键角色。 diff --git a/content/tracks/datalake.md b/content/tracks/datalake.md new file mode 100644 index 000000000..56390572a --- /dev/null +++ b/content/tracks/datalake.md @@ -0,0 +1,9 @@ +--- +title: Data Lake & Data Warehouse +trackchairs: "Lidong Dai, Shaofeng Shi, Zongtang Hu, Jean-Baptiste Onofré, Huaxin Gao" +--- + +Data Lake and Data Warehouse are important solutions for storing and managing data, and they play a crucial role in data management, data analysis, and decision-making. +In ASF, there are various projects about Data Lake and Data Warehouse, for example: +Apache Hive, Apache Hudi, Apache Iceberg, Apache Paimon, Apache Cassandra, Apache HBase, Apache Cloudberry (Incubating) etc. +In this topic, you will get the latest status of data lake and warehouse, best practices the companies use them in the production, and the roadmap of these projects. diff --git a/content/tracks/datalake.zh.md b/content/tracks/datalake.zh.md new file mode 100644 index 000000000..cdc89e897 --- /dev/null +++ b/content/tracks/datalake.zh.md @@ -0,0 +1,9 @@ +--- +title: 数据湖/数据仓库 +trackchairs: "代立冬, 史少锋, 胡宗棠, Jean-Baptiste Onofré, 高华欣" +--- + +数据湖和数据仓库是存储和管理数据的重要解决方案，它们在数据管理、数据分析和决策中发挥着关键作用。 +在ASF，有不少是与数据湖和数据仓库有关的项目，例如： +Apache Hive, Apache Hudi, Apache Iceberg, Apache Paimon, Apache Cassandra, Apache HBase, Apache Cloudberry (Incubating)等。 +在这个主题中，你将得到数据湖和仓库的最新情况，公司在生产中使用它们的最佳做法，以及这些项目的路线图。 diff --git a/content/tracks/dataops.md b/content/tracks/dataops.md new file mode 100644 index 000000000..2e57adcbb --- /dev/null +++ b/content/tracks/dataops.md @@ -0,0 +1,6 @@ +--- +title: DataOps +trackchairs: "William Guo, Lifeng Nie" +--- +Featuring some of the most innovative and cutting-edge projects in the Apache ecosystem. This track brings together leading experts and contributors from Apache DolphinScheduler, Apache Airflow, Apache SeaTunnel, Apache Flume, Apache Sqoop, Apache Griffin, Apache Atlas and other DataOps-related projects to explore the latest advances in data operations, automation, and orchestration. Whether you’re a seasoned data professional or just getting started in the field, this track offers something for everyone, with sessions covering topics such as data pipelines, ETL, orchestration, data qulity, metadata, and more. Join us at ApacheCon for an exciting and informative deep dive into the world of DataOps. + diff --git a/content/tracks/dataops.zh.md b/content/tracks/dataops.zh.md new file mode 100644 index 000000000..c7010b2a1 --- /dev/null +++ b/content/tracks/dataops.zh.md @@ -0,0 +1,6 @@ +--- +title: DataOps +trackchairs: "郭炜,聂励峰" +--- + +本专题聚焦 Apache 生态系统中最具创新性和最前沿的项目。该专题汇聚了来自Apache DolphinScheduler、Apache Airflow、Apache SeaTunnel、Apache Flume、Apache Sqoop、Apache Griffin、Apache Atlas以及其他与DataOps相关的项目的领先专家和贡献者，探索数据操作、自动化和编排的最新进展。无论您是经验丰富的数据专业人员还是刚刚入门的新手，这个专题都为每个人提供了有价值的内容，包括数据管道、ETL、编排、数据质量、元数据等多个主题的会话。在ApacheCon上，加入我们一起深入探索DataOps领域的精彩世界。 diff --git a/content/tracks/datastorage.md b/content/tracks/datastorage.md new file mode 100644 index 000000000..d0c71f602 --- /dev/null +++ b/content/tracks/datastorage.md @@ -0,0 +1,12 @@ +--- +title: Data Storage & Computing +trackchairs: "Gang Li,Nicholas Jiang" +--- + +Big data is an important branch of computer science,Researches and innovations in the big data storage and computing have never stopped. +Big Data is leading and changing various industries and is inseparable from our lives. + +Big Data is also a very important part of ASF. ASF has so many big data storage and computing projects, such as [Apache Hadoop](https://hadoop.apache.org), [Apache Spark](https://spark.apache.org), [Apache HBase](https://hbase.apache.org), [Apache Ozone](https://ozone.apache.org), [Apache CarbonData](https://carbondata.apache.org), [Apache Cassandra](https://cassandra.apache.org), [Apache ZooKeeper](https://zookeeper.apache.org), [Apache Celeborn](https://celeborn.apache.org/) etc. + +In this topic, you will learn the cutting-edge trends of these technologies and the practical experience, principles, architecture analysis and other exciting content from first-line users. + diff --git a/content/tracks/datastorage.zh.md b/content/tracks/datastorage.zh.md new file mode 100644 index 000000000..d85c804a5 --- /dev/null +++ b/content/tracks/datastorage.zh.md @@ -0,0 +1,8 @@ +--- +title: 数据存储与计算 +trackchairs: "李岗，蒋晓峰" +--- + +大数据是计算机科学的一个重要分支,大数据存储和计算领域的研究和创新从未停止。大数据正在深刻的引领和改变着各个行业，已经与我们的生活密不可分。 + +大数据也是 ASF 非常重要的组成部分，ASF 有非常多的大数据存储和计算领域的项目，比如大家熟知的 [Apache Hadoop](https://hadoop.apache.org), [Apache Spark](https://spark.apache.org), [Apache HBase](https://hbase.apache.org), [Apache Ozone](https://ozone.apache.org), [Apache CarbonData](https://carbondata.apache.org), [Apache Cassandra](https://cassandra.apache.org), [Apache ZooKeeper](https://zookeeper.apache.org), [Apache Celeborn](https://celeborn.apache.org/) 等等，在这个主题中，大家会学习到这些技术的前沿趋势和来自一线用户的实践经验、原理、架构分析等精彩内容 diff --git a/content/tracks/drill.md b/content/tracks/drill.md deleted file mode 100755 index 2519f2793..000000000 --- a/content/tracks/drill.md +++ /dev/null @@ -1,3 +0,0 @@ ---- -title: Apache Drill ---- diff --git a/content/tracks/general.md b/content/tracks/general.md old mode 100755 new mode 100644 index d77f990da..dd796675e --- a/content/tracks/general.md +++ b/content/tracks/general.md @@ -1,3 +1,6 @@ --- title: General +trackchairs: "Willem Jiang" --- + +The General track of Community Over Code. \ No newline at end of file diff --git a/content/tracks/general.zh.md b/content/tracks/general.zh.md new file mode 100644 index 000000000..4bd987ed0 --- /dev/null +++ b/content/tracks/general.zh.md @@ -0,0 +1,6 @@ +--- +title: 通用 +trackchairs: "姜宁" +--- + +Apache 顶级项目的实际应用案例。 \ No newline at end of file diff --git a/content/tracks/geospatial.md b/content/tracks/geospatial.md deleted file mode 100755 index ec9391b99..000000000 --- a/content/tracks/geospatial.md +++ /dev/null @@ -1,3 +0,0 @@ ---- -title: Geospatial software ---- diff --git a/content/tracks/graphprocessing.md b/content/tracks/graphprocessing.md deleted file mode 100755 index 06358e7c5..000000000 --- a/content/tracks/graphprocessing.md +++ /dev/null @@ -1,3 +0,0 @@ ---- -title: Graph processing ---- diff --git a/content/tracks/incubator.md b/content/tracks/incubator.md new file mode 100644 index 000000000..3bbec4a43 --- /dev/null +++ b/content/tracks/incubator.md @@ -0,0 +1,8 @@ +--- +title: Incubator +trackchairs: "Calvin Kirs, Justin Mclean" +--- + +The Apache Incubator offers valuable services to projects, referred to as "podlings," seeking entry into the Apache Software Foundation (ASF). This conference track aims to provide insights and guidance on the incubator journey, enabling attendees to gain a deeper understanding of Apache's governance and operational practices. + +Participants will learn about the process of entering the incubator and progressing towards becoming a Top-Level Project (TLP). Additionally, this track will explore strategies for building an active and diverse open-source community, while ensuring compliance with Apache's regulations. diff --git a/content/tracks/incubator.zh.md b/content/tracks/incubator.zh.md new file mode 100644 index 000000000..bd9e69962 --- /dev/null +++ b/content/tracks/incubator.zh.md @@ -0,0 +1,5 @@ +--- +title: Incubator +trackchairs: "郭强，Justin Mclean" +--- +Apache 孵化器为所有希望成为 Apache 软件基金会（ASF）项目提供帮助和服务。本会场，我们将分享与孵化器相关的经验和指导，包括 Apache 软件基金会的治理和运作风格，以及项目如何进入 ASF 孵化器并晋升为顶级项目等。此外，您还将学习到如何建立成熟、活跃和多样化的开源社区，以及如何确保项目的发展符合 Apache 软件基金会的法律法规和 Apache way 的原则等内容。 diff --git a/content/tracks/integration.md b/content/tracks/integration.md deleted file mode 100755 index a22324719..000000000 --- a/content/tracks/integration.md +++ /dev/null @@ -1,3 +0,0 @@ ---- -title: Integration ---- diff --git a/content/tracks/internetofthings.md b/content/tracks/internetofthings.md deleted file mode 100755 index 5c1526cdb..000000000 --- a/content/tracks/internetofthings.md +++ /dev/null @@ -1,3 +0,0 @@ ---- -title: Internet of Things ---- diff --git a/content/tracks/iot.md b/content/tracks/iot.md new file mode 100644 index 000000000..4e48ccbdb --- /dev/null +++ b/content/tracks/iot.md @@ -0,0 +1,13 @@ +--- +title: IoT and IIoT +trackchairs: "Jialin Qiao, Pengcheng Zheng" +--- + +IoT and IIoT focuses on how Apache projects empower connected devices, industrial systems, and AI-driven intelligence at the edge. The focus is on real architectures, cross-project integrations, and production experiences across the IoT stack. + +Topics include: +(1) Embedded platforms and edge AI: real-time OS, device management, and edge-to-cloud synchronization with ML inference for IoT devices. +(2) Data infrastructure and time-series intelligence: time-series databases, stream processing, data pipelines, industrial analytics, and emerging time-series foundation models for forecasting, anomaly detection, and predictive maintenance. + +We welcome talks from engineers and architects sharing concrete designs, implementation tips, and real-world deployment stories, including large-scale device fleet management, reliable data pipelines for ML, cross-project integration patterns, and building end-to-end intelligent IoT systems using Apache technologies. + diff --git a/content/tracks/iot.zh.md b/content/tracks/iot.zh.md new file mode 100644 index 000000000..3466f44d0 --- /dev/null +++ b/content/tracks/iot.zh.md @@ -0,0 +1,10 @@ +--- +title: 工业物联网 +trackchairs: "乔嘉林, 郑鹏程" +--- + +物联网与工业物联网专题聚焦于Apache项目如何为互联设备、工业系统及边缘AI智能提供赋能。重点关注系统架构、跨域集成以及物联网全栈的实践经验。 +主题包括： +（1）嵌入式平台与边缘AI：实时操作系统、设备管理，以及面向物联网设备的边云协同机器学习推理。 +（2）数据基础设施与时序智能：时序数据库、流处理、数据集成、AI分析，以及面向预测、异常检测和预测性维护的新兴时序大模型。 +欢迎工程师和架构师分享具体设计方案、实现技巧及真实部署案例，包括大规模设备集群管理、面向机器学习的可靠数据集成模式，以及基于Apache技术构建端到端智能物联网系统等议题。 \ No newline at end of file diff --git a/content/tracks/karaf.md b/content/tracks/karaf.md deleted file mode 100755 index 4634db515..000000000 --- a/content/tracks/karaf.md +++ /dev/null @@ -1,3 +0,0 @@ ---- -title: Apache Karaf ---- diff --git a/content/tracks/keynote.md b/content/tracks/keynote.md new file mode 100644 index 000000000..f9476fa17 --- /dev/null +++ b/content/tracks/keynote.md @@ -0,0 +1,7 @@ +--- +title: Keynote +trackchairs: "Willem Jiang, Nadia Jiang" +--- +Keynote track is the main track. +we invite many famous people in open source community, +they will share their visions and insights about open source. diff --git a/content/tracks/keynote.zh.md b/content/tracks/keynote.zh.md new file mode 100644 index 000000000..5fdcd71ca --- /dev/null +++ b/content/tracks/keynote.zh.md @@ -0,0 +1,6 @@ +--- +title: Keynote +trackchairs: "姜宁，江波" +--- + +Keynote是主会场，我们邀请了各位开源大咖，来分享他们在开源社区的心得和感悟。 diff --git a/content/tracks/machinelearning.md b/content/tracks/machinelearning.md deleted file mode 100755 index d3b14986c..000000000 --- a/content/tracks/machinelearning.md +++ /dev/null @@ -1,3 +0,0 @@ ---- -title: Machine Learning ---- diff --git a/content/tracks/messaging.md b/content/tracks/messaging.md new file mode 100644 index 000000000..5805b2499 --- /dev/null +++ b/content/tracks/messaging.md @@ -0,0 +1,9 @@ +--- +title: Messaging +trackchairs: "Jia Zhai, Qingshan Lin, Zongtang Hu" +--- + +With the large-scale landing of serverless, IoT and real-time data technology, event-driven architecture and event streaming technology have been more widely applied, making message queue become more and more important infrastructure. Today, a number of excellent messaging projects have emerged in the Apache ecosystem, including +[Apache Pulsar](https://pulsar.apache.org/), [Apache Kafka](https://kafka.apache.org/), [Apache RocketMQ](https://rocketmq.apache.org/), [Apache ActiveMQ](https://activemq.apache.org/), [Apache Inlong](https://inlong.apache.org/), etc, facing new technological trends, each messaging project is also continuing to evolve. + +In this topic, you will learn how different messaging systems make the best technical evolution direction based on their own architectural characteristics, including storage and computing separation, serverless, messaging-streaming integration,and so on. You can also learn how major manufacturers choose the right messaging technology based on their own industry characteristics and business scenarios, and obtain the best practices of messaging technology. diff --git a/content/tracks/messaging.zh.md b/content/tracks/messaging.zh.md new file mode 100644 index 000000000..20687344c --- /dev/null +++ b/content/tracks/messaging.zh.md @@ -0,0 +1,9 @@ +--- +title: 消息系统 +trackchairs: "翟佳, 林清山, 胡宗棠" +--- + +伴随着云原生Serverless、物联网、实时数据技术的规模化落地，事件驱动架构、事件流技术得以更广泛的应用，使得消息队列成为越来越重要的基础设施。今天在Apache生态中已经涌现了多个优秀的消息项目，包括：[Apache Pulsar](https://pulsar.apache.org/), [Apache Kafka](https://kafka.apache.org/) +, [Apache RocketMQ](https://rocketmq.apache.org/), [Apache ActiveMQ](https://activemq.apache.org/), [Apache Inlong](https://inlong.apache.org/) 等等。 + +在这个主题中，你将了解到不同的消息系统如何基于自身的架构特点做出最优的技术演进方向，包括存算分离、Serverless、消息流一体等；也能学习各大厂商如何结合自身的行业特点、业务场景选择合适的消息技术，获得消息技术的最佳实践。 diff --git a/content/tracks/microservice.md b/content/tracks/microservice.md new file mode 100644 index 000000000..167285100 --- /dev/null +++ b/content/tracks/microservice.md @@ -0,0 +1,14 @@ +--- +title: Microservice +trackchairs: "Jun Liu" +--- + +Whether in the era of Cloud Native or AI-led transformation today, microservices remain the cornerstone of underlying systems, helping us build resilient and scalable distributed systems. Join us as we uncover: + +* High-Performance RPC Frameworks – Learn how projects like Apache Dubbo, Apache Thrift, and Apache bRPC are redefining service-to-service communication with speed, efficiency, and interoperability. + +* AI-Driven Microservices – Discover how AI and ML are revolutionizing service orchestration, anomaly detection, and autonomous scaling in microservice ecosystems. + +* The Next-Gen Service Mesh – Beyond traditional sidecars, we explore the rise of Proxyless Service Mesh and how it transforms observability, security, and traffic management in cloud-native environments. + +* Cloud-Native Microservices – From Kubernetes to serverless, we discuss how cloud-native principles optimize microservice architectures for scalability, resilience, and automation. diff --git a/content/tracks/microservice.zh.md b/content/tracks/microservice.zh.md new file mode 100644 index 000000000..48ce599fa --- /dev/null +++ b/content/tracks/microservice.zh.md @@ -0,0 +1,14 @@ +--- +title: 微服务 +trackchairs: "刘军" +--- + +不论是在云原生时代，还是在 AI 引领变革今天，微服务依然是底层系统的基石，它帮助我们构建弹性、可扩展的分布式系统。加入我们，一同了解： + +* 高性能 RPC 框架 – 探讨像 Apache Dubbo、Apache Thrift 和 Apache bRPC 这样的项目如何通过速度、效率和互操作性重新定义服务间通信。 + +* 下一代服务网格 – 超越传统的 Sidecar，我们将探讨 Proxyless Mesh 的崛起，以及它如何在云原生环境中改变可观察性、安全性和流量管理。 + +* 云原生微服务 – 从 Kubernetes 到 Serverless 架构，我们讨论云原生原则如何优化微服务架构的可扩展性、韧性和自动化。 + +* AI 驱动的微服务 – 了解人工智能和机器学习如何在微服务生态系统中变革服务编排、异常检测和自主扩展。 \ No newline at end of file diff --git a/content/tracks/mobile.md b/content/tracks/mobile.md deleted file mode 100755 index a6e5b20a1..000000000 --- a/content/tracks/mobile.md +++ /dev/null @@ -1,3 +0,0 @@ ---- -title: Mobile ---- diff --git a/content/tracks/observability.md b/content/tracks/observability.md old mode 100755 new mode 100644 index 19d027311..ac6fd6493 --- a/content/tracks/observability.md +++ b/content/tracks/observability.md @@ -1,3 +1,6 @@ --- title: Observability +trackchairs: "Huxing Zhang, Sheng Wu" --- + +As an essential technical capability in the cloud-native era, observability—the ability to analyze and interpret a system's operational state—has gained increasing importance. Especially in today's widespread adoption of distributed environments, observability platforms serve as the cornerstone for ensuring software system stability. diff --git a/content/tracks/observability.zh.md b/content/tracks/observability.zh.md new file mode 100644 index 000000000..7dec4602a --- /dev/null +++ b/content/tracks/observability.zh.md @@ -0,0 +1,6 @@ +--- +title: 可观测性 +trackchairs: "张乎兴, 吴晟" +--- + +作为云原生时代必备的技术能力，软件中的可观测性（指分析和解释系统运行状态的能力）越来越受到重视，特别是在分布式环境广泛运用的今天，可观测性平台能力可谓保障软件系统稳定的基石。 diff --git a/content/tracks/ofbiz.md b/content/tracks/ofbiz.md deleted file mode 100755 index 9d5169c4e..000000000 --- a/content/tracks/ofbiz.md +++ /dev/null @@ -1,3 +0,0 @@ ---- -title: OFBiz ---- diff --git a/content/tracks/olap.md b/content/tracks/olap.md new file mode 100644 index 000000000..07751a2ca --- /dev/null +++ b/content/tracks/olap.md @@ -0,0 +1,10 @@ +--- +title: OLAP & Data Analysis +trackchairs: "Mingyu Chen, Zhijing Lu, Dianjin Wang" +--- + +With the advent of the big data era, the application of data analysis and OLAP technology in enterprises is becoming more and more widespread. In order to promote communication and sharing in this field in the industry, we have set up the OLAP and Data Analysis Track. We hope to invite experts and scholars in the industry to share their research results, practical experience, and latest developments in data analysis and OLAP technology. + +In ASF, there are various projects about OLAP and data analysis, for example: Apache Doris, Apache Druid, Apache Kylin, Apache Pinot, Apache Impala, Apache Calcite, Apache Cloudberry (Incubating), etc. + +The topic can cover research progress, application cases, best practices, performance optimization, and other aspects of data analysis and OLAP technology. We believe it will greatly bring great benefits to the attendees. diff --git a/content/tracks/olap.zh.md b/content/tracks/olap.zh.md new file mode 100644 index 000000000..48a4f3ca1 --- /dev/null +++ b/content/tracks/olap.zh.md @@ -0,0 +1,10 @@ +--- +title: OLAP和数据分析 +trackchairs: "陈明雨, 鲁志敬, 王殿进" +--- + +随着大数据时代的到来，数据分析和OLAP技术在企业中的应用也越来越广泛。为了促进业界在这一领域的交流和分享，我们设立了OLAP和数据分析专场。我们希望邀请业内的专家和学者分享他们在数据分析和OLAP技术方面的研究成果、实践经验和最新发展。 + +在ASF，有各种关于OLAP和数据分析的项目，例如： Apache Doris, Apache Druid, Apache Kylin, Apache Pinot, Apache Impala, Apache Calcite, Apache Cloudberry (Incubating) 等。 + +该主题可以涵盖数据分析和OLAP技术的研究进展、应用案例、最佳实践、性能优化等方面。我们相信，这将给与会者带来极大的好处。 diff --git a/content/tracks/rust.md b/content/tracks/rust.md new file mode 100644 index 000000000..c4358809f --- /dev/null +++ b/content/tracks/rust.md @@ -0,0 +1,8 @@ +--- +title: Rust +trackchairs: "tison, Handong Zhang" +--- + +The Rust ecosystem is flourishing. Many ASF projects now offer Rust SDKs, and several new ASF projects are primarily written in Rust, such as Apache OpenDAL and Apache Teaclave, among others. + +The Rust Track aims to share the evolution of Rust projects and integrations within the ASF and broader open-source ecosystem. It will demonstrate how ASF projects provide excellent Rust implementations and illustrate how the incubator helps Rust projects develop healthy communities under the guidance of The Apache Way. diff --git a/content/tracks/rust.zh.md b/content/tracks/rust.zh.md new file mode 100644 index 000000000..bccabf5a3 --- /dev/null +++ b/content/tracks/rust.zh.md @@ -0,0 +1,8 @@ +--- +title: Rust +trackchairs: "陈梓立, 张汉东" +--- + +Rust 生态系统正在蓬勃发展。许多 ASF 项目都提供了 Rust SDK，不少新的 ASF 项目主要用 Rust 编写，例如 Apache OpenDAL 和 Apache Teaclave 等等。 + +Rust 专场旨在分享 ASF 内部及开源生态中 Rust 项目和集成的演变历程，展示 ASF 项目如何提供良好的 Rust 实践，并说明孵化器如何帮助 Rust 项目在 The Apache Way 的指导下健康地发展社区。 diff --git a/content/tracks/streaming.md b/content/tracks/streaming.md old mode 100755 new mode 100644 index b11ee8533..9fbde7a9c --- a/content/tracks/streaming.md +++ b/content/tracks/streaming.md @@ -1,3 +1,7 @@ --- title: Streaming +trackchairs: "Yu Li, Xin Wang" --- +Streaming data processing is a big deal in big data these days, businesses crave ever-more timely insights into their data, and what was once a ‘batch’ mindset is quickly being replaced with stream processing. More and more companies, small and large, are rethinking their architecture with real-time context at the forefront, and starting to build their streaming platforms with powerful open source engines such as [Apache Flink](https://flink.apache.org/), [Apache Spark](https://spark.apache.org/), [Apache Kafka](https://kafka.apache.org/), [Apache Pulsar](https://pulsar.apache.org/), [Apache Storm](https://storm.apache.org/), [Apache StreamPark](https://streampark.apache.org/), [Apache Paimon](https://paimon.apache.org) etc. + +In this topic, you will not only learn about the practical experience of first-line users in applying these Apache projects to their in-production environment, but also the latest developments in the ecology of these Apache projects, and visions on where streaming technology is heading in the future. diff --git a/content/tracks/streaming.zh.md b/content/tracks/streaming.zh.md new file mode 100644 index 000000000..187f9ea62 --- /dev/null +++ b/content/tracks/streaming.zh.md @@ -0,0 +1,7 @@ +--- +title: 流处理 +trackchairs: "李钰, 王鑫" +--- +流式数据处理是当今大数据领域的趋势，很多企业渴望更及时地洞察自己的数据，而曾经的“批处理”思维正迅速被流式处理所取代。越来越多的公司，无论大小，都在重新思考技术架构时把实时性作为第一考量，并开始用强大的开源引擎如[Apache Flink](https://flink.apache.org/), [Apache Spark](https://spark.apache.org/), [Apache Kafka](https://kafka.apache.org/), [Apache Pulsar](https://pulsar.apache.org/), [Apache Storm](https://storm.apache.org/), [Apache StreamPark](https://streampark.apache.org/), [Apache Paimon](https://paimon.apache.org) 等构建自己的实时计算平台。 + +在该主题中，您将了解到一线大厂把这些Apache项目应用到其生产环境中的实际经验，以及这些Apache项目生态的最新发展和流计算技术未来的发展方向。 diff --git a/content/tracks/tomcat.md b/content/tracks/tomcat.md deleted file mode 100755 index 02843faaa..000000000 --- a/content/tracks/tomcat.md +++ /dev/null @@ -1,3 +0,0 @@ ---- -title: Apache Tomcat ---- diff --git a/content/tracks/webserver.md b/content/tracks/webserver.md new file mode 100644 index 000000000..83410f21f --- /dev/null +++ b/content/tracks/webserver.md @@ -0,0 +1,11 @@ +--- +title: Web Application & Framework +trackchairs: "Huxing Zhang, Han Li" +--- + +With the growing demands of modern web applications, building efficient, scalable, and secure web systems has become a core objective of technological development. + +Within the Apache ecosystem, numerous outstanding projects have emerged, ranging from classic web servers (such as Tomcat, HTTP Server) to essential development tools (like Apache Commons, FreeMarker, Echarts), as well as frameworks specializing in enterprise-level integration and security (e.g., Camel, Shiro). These projects play critical roles in modern web architectures. + +Here, you will gain in-depth insights into the latest developments of these projects, learn best practices for production environments, and explore their future directions. + diff --git a/content/tracks/webserver.zh.md b/content/tracks/webserver.zh.md new file mode 100644 index 000000000..d307ae232 --- /dev/null +++ b/content/tracks/webserver.zh.md @@ -0,0 +1,10 @@ +--- +title: Web 应用与框架 +trackchairs: "张乎兴, 李晗" +--- + +随着现代 Web 应用需求的不断增长，构建高效、可扩展且安全的 Web 系统已成为技术发展的核心目标。 + +在 Apache 生态系统中，涌现了众多优秀项目，从经典的 Web 服务器（如 Tomcat、HTTP Server）到基础开发工具（如 Apache Commons、FreeMarker、Echarts），再到专注于企业级集成和安全的框架（如 Camel、Shiro）。这些项目在现代 Web 架构中扮演着至关重要的角色。 + +在这里，您将深入了解这些项目的最新进展，学习生产环境中的最佳实践，并探讨它们未来的发展方向。 diff --git a/content/venue_travel/travel.md b/content/venue_travel/travel.md new file mode 100644 index 000000000..36b5b6e5a --- /dev/null +++ b/content/venue_travel/travel.md @@ -0,0 +1,14 @@ +--- +title: Business Travel +--- + +### Conference Hotel + +Park Plaza Beijing Science Park Hotel is 4.6km from the venue, approximately 20 minutes by taxi. + +### Nearby Hotels (within 3km) +- Beijing Post and Telecommunications Conference Center +- Langlizi Hotel (Suzhou Street, Zhongguancun, Beijing) +- Ji Hotel (Suzhou Street, Zhongguancun, Beijing) +- Crowne Plaza Beijing Zhongguancun +- Century Huatian Hotel \ No newline at end of file diff --git a/content/venue_travel/travel.zh.md b/content/venue_travel/travel.zh.md new file mode 100644 index 000000000..863d10f14 --- /dev/null +++ b/content/venue_travel/travel.zh.md @@ -0,0 +1,15 @@ +--- +title: 差旅住宿 +--- + +### 会议协议酒店 + +北京丽亭华苑酒店，距离会场直线距离 4.6km，打车约 20 分钟抵达会议酒店。 + +### 周边酒店（直线距离 3km 内） + +- 北京邮电会议中心 +- 朗丽兹酒店（北京中关村苏州街店） +- 全季酒店（北京中关村苏州街店） +- 北京中关村皇冠假日酒店 +- 世纪华天大酒店 \ No newline at end of file diff --git a/content/venue_travel/venue.md b/content/venue_travel/venue.md new file mode 100644 index 000000000..9493376b0 --- /dev/null +++ b/content/venue_travel/venue.md @@ -0,0 +1,34 @@ +--- +title: Venue + Travel +weight: 20 +--- + +### Venue + +This year's conference will be held at the Convention Center of Zhongguancun National Independent Innovation Demonstration Zone (located at No. 2 Xinjian Gongmen Road, Haidian District, Beijing). + +Nestled in the heart of Haidian District, Beijing, the Convention Center of Zhongguancun National Independent Innovation Demonstration Zone stands adjacent to the time-honored cultural scenic area of "Three Hills and Five Gardens" and the thriving Zhongguancun Science City. It serves as a confluence point of technology and culture, as well as a brilliant stage for innovation and exchange. Boasting a vast total construction area and a scientifically designed interior layout, the convention center features multiple functional meeting rooms that can accommodate meeting events of various scales and needs, ranging from small seminars for several dozen participants to large-scale summits for over a thousand attendees. Thanks to its unique architectural design, state-of-the-art conference facilities, and exceptional geographical location, the venue has become the premier choice for hosting a wide array of high-end conferences, forums, exhibitions, and business events both domestically and internationally, setting a new trend for the integration of technological innovation and cultural exchange. + +
+
+ +
+
+ +
+
+ +
+ +
+ +
+
+ +### Getting There + +**Taking a taxi**: Search for "Zhongguancun National Innovation Demonstration Zone Conference Center". + +**Taking the subway**: Take Metro Line 16 to the "Wanquanheqiao" Station. Exit from Exit A, then walk for 700 meters (about 11 minutes) to reach the park. + +**Location**: [Zhongguancun National Zizhu Innovative Demonstration Area - Conference Center](https://maps.apple.com/?address=%E4%B8%AD%E5%9B%BD%E5%8C%97%E4%BA%AC%E5%B8%82%E6%B5%B7%E6%B7%80%E5%8C%BA%E6%96%B0%E5%BB%BA%E5%AE%AB%E9%97%A8%E8%B7%AF2%E5%8F%B7&auid=1118368620452725&ll=39.989483,116.293350&lsp=57879&q=%E4%B8%AD%E5%85%B3%E6%9D%91%E5%9B%BD%E5%AE%B6%E8%87%AA%E4%B8%BB%E5%88%9B%E6%96%B0%E7%A4%BA%E8%8C%83%E5%8C%BA-%E4%BC%9A%E8%AE%AE%E4%B8%AD%E5%BF%83&_ext=EiQpLp67o0X9Q0AxmD0PjgUSXUA5/XUt2wgAREBBqUGSuYYTXUA%3D) diff --git a/content/venue_travel/venue.zh.md b/content/venue_travel/venue.zh.md new file mode 100644 index 000000000..593790f21 --- /dev/null +++ b/content/venue_travel/venue.zh.md @@ -0,0 +1,36 @@ +--- +title: 会场 + 旅行 +weight: 20 +--- + +### Venue 会议场地 + +本届大会将在北京市海淀区中关村国家自主创新示范区展示交易中心-会议中心（北京市海淀区新建宫门路2号中关村展示中心-5号门）举办。 + +关村国家自主创新示范区展示交易中心-会议中心，坐落于北京市海淀区的心脏地带，紧邻历史悠久的三山五园文化景区与蓬勃发展的中关村科学城，是科技与文化的交汇点，也是创新与交流的璀璨舞台。会议中心总建筑面积宏大，内部空间布局科学合理，拥有多间多功能会议室，从容纳数十人的小型研讨会到上千人的大型峰会，均能满足不同规模与需求的会议活动。场地凭借其独特的建筑设计、先进的会议设施以及卓越的地理位置，成为了国内外各类高端会议、论坛、展览及商务活动的首选之地，引领着科技创新与文化交融的新风尚。 + +
+
+ +
+
+ +
+
+ +
+ +
+ +
+
+ +### Getting There + +**打车**：搜索「中关村国家自主创新示范区展示交易中心-会议中心（北京市海淀区新建宫门路2号中关村展示中心-5号门）」 + +**地铁**：搭乘地铁 16 号线至「万泉河桥」站，出 A 口后步行 700 米、11 分钟到达园区。 + +**地点**：[中关村国家自主创新示范区展示交易中心-会议中心](https://maps.apple.com/?address=%E4%B8%AD%E5%9B%BD%E5%8C%97%E4%BA%AC%E5%B8%82%E6%B5%B7%E6%B7%80%E5%8C%BA%E6%96%B0%E5%BB%BA%E5%AE%AB%E9%97%A8%E8%B7%AF2%E5%8F%B7&auid=1118368620452725&ll=39.989483,116.293350&lsp=57879&q=%E4%B8%AD%E5%85%B3%E6%9D%91%E5%9B%BD%E5%AE%B6%E8%87%AA%E4%B8%BB%E5%88%9B%E6%96%B0%E7%A4%BA%E8%8C%83%E5%8C%BA-%E4%BC%9A%E8%AE%AE%E4%B8%AD%E5%BF%83&_ext=EiQpLp67o0X9Q0AxmD0PjgUSXUA5/XUt2wgAREBBqUGSuYYTXUA%3D) + + diff --git a/data/guides.yaml b/data/guides.yaml new file mode 100644 index 000000000..388c3fc58 --- /dev/null +++ b/data/guides.yaml @@ -0,0 +1,17 @@ +- name: visa_letter_request + url: /guide/visa_letter_request.html + +- name: presentation_slides_template + url: /guide/presentation_slides_template.html + +- name: convince_your_boss + url: /guide/convince_your_boss.html + +- name: code_of_conduct + url: /guide/code_of_conduct.html + +- name: terms_condition + url: /guide/terms.html + +- name: accessibility + url: /guide/accessibility.html diff --git a/data/rooms.yaml b/data/rooms.yaml new file mode 100644 index 000000000..444bcee96 --- /dev/null +++ b/data/rooms.yaml @@ -0,0 +1,25 @@ +# Column order of the session grid on /sessions, per language. +# Room names must match the `room` front matter of content/sessions/*.md. +# Rooms that are missing here are appended automatically (ordered by the +# first session time, then by track), so the schedule never breaks when the +# venue data changes. +en: + - "MainRoom - YiHe Hall" + - "YuanMing Hall" + - "WanChun Hall" + - "JingMing Hall" + - "JingYi Hall" + - "Mtn WanShou Hall" + - "Mtn BaiWang Hall" + - "Mtn YuQuan Hall" + - "Mtn Yang Hall" +zh: + - "主会场 - 颐和厅" + - "圆明厅" + - "万春厅" + - "静明厅" + - "静宜厅" + - "万寿山会议室" + - "百望山会议室" + - "玉泉山会议室" + - "阳山会议室" diff --git a/docs/categories/index.xml b/docs/categories/index.xml deleted file mode 100644 index 9cca0c849..000000000 --- a/docs/categories/index.xml +++ /dev/null @@ -1,10 +0,0 @@ - - - - Categories on ApacheCon at Home 2021 - https://elek.github.io/acah2021/categories.html - Recent content in Categories on ApacheCon at Home 2021 - Hugo -- gohugo.io - en-us - - diff --git a/docs/cfp.html b/docs/cfp.html deleted file mode 100644 index 815104777..000000000 --- a/docs/cfp.html +++ /dev/null @@ -1,125 +0,0 @@ - - - - - - - - ApacheCon @Home - - - - - - - - - - - - - - - - - - - -
- - - - - - -
- - -
-
-
-

Call for Presentations

-
- -

The call for presentations is now open! Submit proposals by Monday, May 3rd, 2021 8:00 AM (America/New_York time - UTC -5). Please do not wait until the last minute.

-

Submit your talk proposal now!

-

We are looking for presentations about anything relating to Apache Software Foundation projects, open source governance, community, and software development.

-

In particular, this year we are building content tracks around the following specific topics/projects:

-
    -
  • API/Microservice
    • -
  • Cassandra
    • -
  • Content Delivery
    • -
  • Community
    • -
  • Drill
    • -
  • Geospatial
    • -
  • Incubating
    • -
  • IoT
    • -
  • Observability
    • -
  • Royale
    • -
  • Tomcat
    • -
  • Big Data
    • -
  • Community
    • -
  • Content Management
    • -
  • Fineract/Fintech
    • -
  • Groovy
    • -
  • Integration
    • -
  • Karaf
    • -
  • RDF/Linked Data
    • -
  • Search
    • -
  • Social Data
    • -
-

If your proposed presentation falls into one of these categories, please select that topic in the CfP entry form. Or select General if it’s related to some other topic or project area.

-

Submit your talk proposal now!

- - -
-
-
- - - - - -
- - - - - - diff --git a/docs/conduct.html b/docs/conduct.html deleted file mode 100644 index 23fffd070..000000000 --- a/docs/conduct.html +++ /dev/null @@ -1,103 +0,0 @@ - - - - - - - - ApacheCon @Home - - - - - - - - - - - - - - - - - - - -
- - - - - - -
- - -
-
-
-

ApacheCon Code of Conduct

-
- -

At the Apache Software Foundation, all online interactions are governed by the ASF Code of Conduct, while in-person events are subject to the Anti-Harassment Policy.

-

ApacheCon is dedicated to providing a harassment-free experience for everyone. We do not tolerate harassment of participants in any form. Participants violating these rules may be sanctioned or expelled without a refund, at the discretion of the event organizers.

-

Harassment includes offensive verbal comments, deliberate intimidation, stalking, following, unwanted photography or recording, sustained disruption of talks or other events, inappropriate physical contact, and unwelcome sexual attention. Participants asked to stop any harassing behaviour are expected to comply immediately.

-

Sexual language and imagery will not be tolerated in any event venue, including talks. Exhibitors should also refrain from using sexualised images, activities, or other material. Booth staff (including volunteers) should not use sexualised clothing/uniforms/costumes, or otherwise create a sexualised environment.

-

If a participant engages in harassing behaviour, the organizers may take any action they deem appropriate, including warning the offender or expulsion from the event without a refund. We expect participants to follow these rules at all event venues and related social events.

-

If you are being harassed, notice that someone else is being harassed, or have any other concerns, please contact a member of the event team immediately. The team can be found at the registration desk.

-

You can contact the event team via email at planners@apachecon.com, or via the live chat feature on the event website.

- - -
-
-
- - - - - -
- - - - - - diff --git a/docs/css/font-awesome.min.css b/docs/css/font-awesome.min.css deleted file mode 100644 index f61c62a82..000000000 --- a/docs/css/font-awesome.min.css +++ /dev/null @@ -1,4 +0,0 @@ -/*! - * Font Awesome 4.0.3 by @davegandy - http://fontawesome.io - @fontawesome - * License - http://fontawesome.io/license (Font: SIL OFL 1.1, CSS: MIT License) - */@font-face{font-family:'FontAwesome';src:url('../fonts/fontawesome-webfont.eot?v=4.0.3');src:url('../fonts/fontawesome-webfont.eot?#iefix&v=4.0.3') format('embedded-opentype'),url('../fonts/fontawesome-webfont.woff?v=4.0.3') format('woff'),url('../fonts/fontawesome-webfont.ttf?v=4.0.3') format('truetype'),url('../fonts/fontawesome-webfont.svg?v=4.0.3#fontawesomeregular') format('svg');font-weight:normal;font-style:normal}.fa{display:inline-block;font-family:FontAwesome;font-style:normal;font-weight:normal;line-height:1;-webkit-font-smoothing:antialiased;-moz-osx-font-smoothing:grayscale}.fa-lg{font-size:1.3333333333333333em;line-height:.75em;vertical-align:-15%}.fa-2x{font-size:2em}.fa-3x{font-size:3em}.fa-4x{font-size:4em}.fa-5x{font-size:5em}.fa-fw{width:1.2857142857142858em;text-align:center}.fa-ul{padding-left:0;margin-left:2.142857142857143em;list-style-type:none}.fa-ul>li{position:relative}.fa-li{position:absolute;left:-2.142857142857143em;width:2.142857142857143em;top:.14285714285714285em;text-align:center}.fa-li.fa-lg{left:-1.8571428571428572em}.fa-border{padding:.2em .25em .15em;border:solid .08em #eee;border-radius:.1em}.pull-right{float:right}.pull-left{float:left}.fa.pull-left{margin-right:.3em}.fa.pull-right{margin-left:.3em}.fa-spin{-webkit-animation:spin 2s infinite linear;-moz-animation:spin 2s infinite linear;-o-animation:spin 2s infinite linear;animation:spin 2s infinite linear}@-moz-keyframes spin{0%{-moz-transform:rotate(0deg)}100%{-moz-transform:rotate(359deg)}}@-webkit-keyframes spin{0%{-webkit-transform:rotate(0deg)}100%{-webkit-transform:rotate(359deg)}}@-o-keyframes spin{0%{-o-transform:rotate(0deg)}100%{-o-transform:rotate(359deg)}}@-ms-keyframes spin{0%{-ms-transform:rotate(0deg)}100%{-ms-transform:rotate(359deg)}}@keyframes spin{0%{transform:rotate(0deg)}100%{transform:rotate(359deg)}}.fa-rotate-90{filter:progid:DXImageTransform.Microsoft.BasicImage(rotation=1);-webkit-transform:rotate(90deg);-moz-transform:rotate(90deg);-ms-transform:rotate(90deg);-o-transform:rotate(90deg);transform:rotate(90deg)}.fa-rotate-180{filter:progid:DXImageTransform.Microsoft.BasicImage(rotation=2);-webkit-transform:rotate(180deg);-moz-transform:rotate(180deg);-ms-transform:rotate(180deg);-o-transform:rotate(180deg);transform:rotate(180deg)}.fa-rotate-270{filter:progid:DXImageTransform.Microsoft.BasicImage(rotation=3);-webkit-transform:rotate(270deg);-moz-transform:rotate(270deg);-ms-transform:rotate(270deg);-o-transform:rotate(270deg);transform:rotate(270deg)}.fa-flip-horizontal{filter:progid:DXImageTransform.Microsoft.BasicImage(rotation=0,mirror=1);-webkit-transform:scale(-1,1);-moz-transform:scale(-1,1);-ms-transform:scale(-1,1);-o-transform:scale(-1,1);transform:scale(-1,1)}.fa-flip-vertical{filter:progid:DXImageTransform.Microsoft.BasicImage(rotation=2,mirror=1);-webkit-transform:scale(1,-1);-moz-transform:scale(1,-1);-ms-transform:scale(1,-1);-o-transform:scale(1,-1);transform:scale(1,-1)}.fa-stack{position:relative;display:inline-block;width:2em;height:2em;line-height:2em;vertical-align:middle}.fa-stack-1x,.fa-stack-2x{position:absolute;left:0;width:100%;text-align:center}.fa-stack-1x{line-height:inherit}.fa-stack-2x{font-size:2em}.fa-inverse{color:#ff0011}.fa-glass:before{content:"\f000"}.fa-music:before{content:"\f001"}.fa-search:before{content:"\f002"}.fa-envelope-o:before{content:"\f003"}.fa-heart:before{content:"\f004"}.fa-star:before{content:"\f005"}.fa-star-o:before{content:"\f006"}.fa-user:before{content:"\f007"}.fa-film:before{content:"\f008"}.fa-th-large:before{content:"\f009"}.fa-th:before{content:"\f00a"}.fa-th-list:before{content:"\f00b"}.fa-check:before{content:"\f00c"}.fa-times:before{content:"\f00d"}.fa-search-plus:before{content:"\f00e"}.fa-search-minus:before{content:"\f010"}.fa-power-off:before{content:"\f011"}.fa-signal:before{content:"\f012"}.fa-gear:before,.fa-cog:before{content:"\f013"}.fa-trash-o:before{content:"\f014"}.fa-home:before{content:"\f015"}.fa-file-o:before{content:"\f016"}.fa-clock-o:before{content:"\f017"}.fa-road:before{content:"\f018"}.fa-download:before{content:"\f019"}.fa-arrow-circle-o-down:before{content:"\f01a"}.fa-arrow-circle-o-up:before{content:"\f01b"}.fa-inbox:before{content:"\f01c"}.fa-play-circle-o:before{content:"\f01d"}.fa-rotate-right:before,.fa-repeat:before{content:"\f01e"}.fa-refresh:before{content:"\f021"}.fa-list-alt:before{content:"\f022"}.fa-lock:before{content:"\f023"}.fa-flag:before{content:"\f024"}.fa-headphones:before{content:"\f025"}.fa-volume-off:before{content:"\f026"}.fa-volume-down:before{content:"\f027"}.fa-volume-up:before{content:"\f028"}.fa-qrcode:before{content:"\f029"}.fa-barcode:before{content:"\f02a"}.fa-tag:before{content:"\f02b"}.fa-tags:before{content:"\f02c"}.fa-book:before{content:"\f02d"}.fa-bookmark:before{content:"\f02e"}.fa-print:before{content:"\f02f"}.fa-camera:before{content:"\f030"}.fa-font:before{content:"\f031"}.fa-bold:before{content:"\f032"}.fa-italic:before{content:"\f033"}.fa-text-height:before{content:"\f034"}.fa-text-width:before{content:"\f035"}.fa-align-left:before{content:"\f036"}.fa-align-center:before{content:"\f037"}.fa-align-right:before{content:"\f038"}.fa-align-justify:before{content:"\f039"}.fa-list:before{content:"\f03a"}.fa-dedent:before,.fa-outdent:before{content:"\f03b"}.fa-indent:before{content:"\f03c"}.fa-video-camera:before{content:"\f03d"}.fa-picture-o:before{content:"\f03e"}.fa-pencil:before{content:"\f040"}.fa-map-marker:before{content:"\f041"}.fa-adjust:before{content:"\f042"}.fa-tint:before{content:"\f043"}.fa-edit:before,.fa-pencil-square-o:before{content:"\f044"}.fa-share-square-o:before{content:"\f045"}.fa-check-square-o:before{content:"\f046"}.fa-arrows:before{content:"\f047"}.fa-step-backward:before{content:"\f048"}.fa-fast-backward:before{content:"\f049"}.fa-backward:before{content:"\f04a"}.fa-play:before{content:"\f04b"}.fa-pause:before{content:"\f04c"}.fa-stop:before{content:"\f04d"}.fa-forward:before{content:"\f04e"}.fa-fast-forward:before{content:"\f050"}.fa-step-forward:before{content:"\f051"}.fa-eject:before{content:"\f052"}.fa-chevron-left:before{content:"\f053"}.fa-chevron-right:before{content:"\f054"}.fa-plus-circle:before{content:"\f055"}.fa-minus-circle:before{content:"\f056"}.fa-times-circle:before{content:"\f057"}.fa-check-circle:before{content:"\f058"}.fa-question-circle:before{content:"\f059"}.fa-info-circle:before{content:"\f05a"}.fa-crosshairs:before{content:"\f05b"}.fa-times-circle-o:before{content:"\f05c"}.fa-check-circle-o:before{content:"\f05d"}.fa-ban:before{content:"\f05e"}.fa-arrow-left:before{content:"\f060"}.fa-arrow-right:before{content:"\f061"}.fa-arrow-up:before{content:"\f062"}.fa-arrow-down:before{content:"\f063"}.fa-mail-forward:before,.fa-share:before{content:"\f064"}.fa-expand:before{content:"\f065"}.fa-compress:before{content:"\f066"}.fa-plus:before{content:"\f067"}.fa-minus:before{content:"\f068"}.fa-asterisk:before{content:"\f069"}.fa-exclamation-circle:before{content:"\f06a"}.fa-gift:before{content:"\f06b"}.fa-leaf:before{content:"\f06c"}.fa-fire:before{content:"\f06d"}.fa-eye:before{content:"\f06e"}.fa-eye-slash:before{content:"\f070"}.fa-warning:before,.fa-exclamation-triangle:before{content:"\f071"}.fa-plane:before{content:"\f072"}.fa-calendar:before{content:"\f073"}.fa-random:before{content:"\f074"}.fa-comment:before{content:"\f075"}.fa-magnet:before{content:"\f076"}.fa-chevron-up:before{content:"\f077"}.fa-chevron-down:before{content:"\f078"}.fa-retweet:before{content:"\f079"}.fa-shopping-cart:before{content:"\f07a"}.fa-folder:before{content:"\f07b"}.fa-folder-open:before{content:"\f07c"}.fa-arrows-v:before{content:"\f07d"}.fa-arrows-h:before{content:"\f07e"}.fa-bar-chart-o:before{content:"\f080"}.fa-twitter-square:before{content:"\f081"}.fa-facebook-square:before{content:"\f082"}.fa-camera-retro:before{content:"\f083"}.fa-key:before{content:"\f084"}.fa-gears:before,.fa-cogs:before{content:"\f085"}.fa-comments:before{content:"\f086"}.fa-thumbs-o-up:before{content:"\f087"}.fa-thumbs-o-down:before{content:"\f088"}.fa-star-half:before{content:"\f089"}.fa-heart-o:before{content:"\f08a"}.fa-sign-out:before{content:"\f08b"}.fa-linkedin-square:before{content:"\f08c"}.fa-thumb-tack:before{content:"\f08d"}.fa-external-link:before{content:"\f08e"}.fa-sign-in:before{content:"\f090"}.fa-trophy:before{content:"\f091"}.fa-github-square:before{content:"\f092"}.fa-upload:before{content:"\f093"}.fa-lemon-o:before{content:"\f094"}.fa-phone:before{content:"\f095"}.fa-square-o:before{content:"\f096"}.fa-bookmark-o:before{content:"\f097"}.fa-phone-square:before{content:"\f098"}.fa-twitter:before{content:"\f099"}.fa-facebook:before{content:"\f09a"}.fa-github:before{content:"\f09b"}.fa-unlock:before{content:"\f09c"}.fa-credit-card:before{content:"\f09d"}.fa-rss:before{content:"\f09e"}.fa-hdd-o:before{content:"\f0a0"}.fa-bullhorn:before{content:"\f0a1"}.fa-bell:before{content:"\f0f3"}.fa-certificate:before{content:"\f0a3"}.fa-hand-o-right:before{content:"\f0a4"}.fa-hand-o-left:before{content:"\f0a5"}.fa-hand-o-up:before{content:"\f0a6"}.fa-hand-o-down:before{content:"\f0a7"}.fa-arrow-circle-left:before{content:"\f0a8"}.fa-arrow-circle-right:before{content:"\f0a9"}.fa-arrow-circle-up:before{content:"\f0aa"}.fa-arrow-circle-down:before{content:"\f0ab"}.fa-globe:before{content:"\f0ac"}.fa-wrench:before{content:"\f0ad"}.fa-tasks:before{content:"\f0ae"}.fa-filter:before{content:"\f0b0"}.fa-briefcase:before{content:"\f0b1"}.fa-arrows-alt:before{content:"\f0b2"}.fa-group:before,.fa-users:before{content:"\f0c0"}.fa-chain:before,.fa-link:before{content:"\f0c1"}.fa-cloud:before{content:"\f0c2"}.fa-flask:before{content:"\f0c3"}.fa-cut:before,.fa-scissors:before{content:"\f0c4"}.fa-copy:before,.fa-files-o:before{content:"\f0c5"}.fa-paperclip:before{content:"\f0c6"}.fa-save:before,.fa-floppy-o:before{content:"\f0c7"}.fa-square:before{content:"\f0c8"}.fa-bars:before{content:"\f0c9"}.fa-list-ul:before{content:"\f0ca"}.fa-list-ol:before{content:"\f0cb"}.fa-strikethrough:before{content:"\f0cc"}.fa-underline:before{content:"\f0cd"}.fa-table:before{content:"\f0ce"}.fa-magic:before{content:"\f0d0"}.fa-truck:before{content:"\f0d1"}.fa-pinterest:before{content:"\f0d2"}.fa-pinterest-square:before{content:"\f0d3"}.fa-google-plus-square:before{content:"\f0d4"}.fa-google-plus:before{content:"\f0d5"}.fa-money:before{content:"\f0d6"}.fa-caret-down:before{content:"\f0d7"}.fa-caret-up:before{content:"\f0d8"}.fa-caret-left:before{content:"\f0d9"}.fa-caret-right:before{content:"\f0da"}.fa-columns:before{content:"\f0db"}.fa-unsorted:before,.fa-sort:before{content:"\f0dc"}.fa-sort-down:before,.fa-sort-asc:before{content:"\f0dd"}.fa-sort-up:before,.fa-sort-desc:before{content:"\f0de"}.fa-envelope:before{content:"\f0e0"}.fa-linkedin:before{content:"\f0e1"}.fa-rotate-left:before,.fa-undo:before{content:"\f0e2"}.fa-legal:before,.fa-gavel:before{content:"\f0e3"}.fa-dashboard:before,.fa-tachometer:before{content:"\f0e4"}.fa-comment-o:before{content:"\f0e5"}.fa-comments-o:before{content:"\f0e6"}.fa-flash:before,.fa-bolt:before{content:"\f0e7"}.fa-sitemap:before{content:"\f0e8"}.fa-umbrella:before{content:"\f0e9"}.fa-paste:before,.fa-clipboard:before{content:"\f0ea"}.fa-lightbulb-o:before{content:"\f0eb"}.fa-exchange:before{content:"\f0ec"}.fa-cloud-download:before{content:"\f0ed"}.fa-cloud-upload:before{content:"\f0ee"}.fa-user-md:before{content:"\f0f0"}.fa-stethoscope:before{content:"\f0f1"}.fa-suitcase:before{content:"\f0f2"}.fa-bell-o:before{content:"\f0a2"}.fa-coffee:before{content:"\f0f4"}.fa-cutlery:before{content:"\f0f5"}.fa-file-text-o:before{content:"\f0f6"}.fa-building-o:before{content:"\f0f7"}.fa-hospital-o:before{content:"\f0f8"}.fa-ambulance:before{content:"\f0f9"}.fa-medkit:before{content:"\f0fa"}.fa-fighter-jet:before{content:"\f0fb"}.fa-beer:before{content:"\f0fc"}.fa-h-square:before{content:"\f0fd"}.fa-plus-square:before{content:"\f0fe"}.fa-angle-double-left:before{content:"\f100"}.fa-angle-double-right:before{content:"\f101"}.fa-angle-double-up:before{content:"\f102"}.fa-angle-double-down:before{content:"\f103"}.fa-angle-left:before{content:"\f104"}.fa-angle-right:before{content:"\f105"}.fa-angle-up:before{content:"\f106"}.fa-angle-down:before{content:"\f107"}.fa-desktop:before{content:"\f108"}.fa-laptop:before{content:"\f109"}.fa-tablet:before{content:"\f10a"}.fa-mobile-phone:before,.fa-mobile:before{content:"\f10b"}.fa-circle-o:before{content:"\f10c"}.fa-quote-left:before{content:"\f10d"}.fa-quote-right:before{content:"\f10e"}.fa-spinner:before{content:"\f110"}.fa-circle:before{content:"\f111"}.fa-mail-reply:before,.fa-reply:before{content:"\f112"}.fa-github-alt:before{content:"\f113"}.fa-folder-o:before{content:"\f114"}.fa-folder-open-o:before{content:"\f115"}.fa-smile-o:before{content:"\f118"}.fa-frown-o:before{content:"\f119"}.fa-meh-o:before{content:"\f11a"}.fa-gamepad:before{content:"\f11b"}.fa-keyboard-o:before{content:"\f11c"}.fa-flag-o:before{content:"\f11d"}.fa-flag-checkered:before{content:"\f11e"}.fa-terminal:before{content:"\f120"}.fa-code:before{content:"\f121"}.fa-reply-all:before{content:"\f122"}.fa-mail-reply-all:before{content:"\f122"}.fa-star-half-empty:before,.fa-star-half-full:before,.fa-star-half-o:before{content:"\f123"}.fa-location-arrow:before{content:"\f124"}.fa-crop:before{content:"\f125"}.fa-code-fork:before{content:"\f126"}.fa-unlink:before,.fa-chain-broken:before{content:"\f127"}.fa-question:before{content:"\f128"}.fa-info:before{content:"\f129"}.fa-exclamation:before{content:"\f12a"}.fa-superscript:before{content:"\f12b"}.fa-subscript:before{content:"\f12c"}.fa-eraser:before{content:"\f12d"}.fa-puzzle-piece:before{content:"\f12e"}.fa-microphone:before{content:"\f130"}.fa-microphone-slash:before{content:"\f131"}.fa-shield:before{content:"\f132"}.fa-calendar-o:before{content:"\f133"}.fa-fire-extinguisher:before{content:"\f134"}.fa-rocket:before{content:"\f135"}.fa-maxcdn:before{content:"\f136"}.fa-chevron-circle-left:before{content:"\f137"}.fa-chevron-circle-right:before{content:"\f138"}.fa-chevron-circle-up:before{content:"\f139"}.fa-chevron-circle-down:before{content:"\f13a"}.fa-html5:before{content:"\f13b"}.fa-css3:before{content:"\f13c"}.fa-anchor:before{content:"\f13d"}.fa-unlock-alt:before{content:"\f13e"}.fa-bullseye:before{content:"\f140"}.fa-ellipsis-h:before{content:"\f141"}.fa-ellipsis-v:before{content:"\f142"}.fa-rss-square:before{content:"\f143"}.fa-play-circle:before{content:"\f144"}.fa-ticket:before{content:"\f145"}.fa-minus-square:before{content:"\f146"}.fa-minus-square-o:before{content:"\f147"}.fa-level-up:before{content:"\f148"}.fa-level-down:before{content:"\f149"}.fa-check-square:before{content:"\f14a"}.fa-pencil-square:before{content:"\f14b"}.fa-external-link-square:before{content:"\f14c"}.fa-share-square:before{content:"\f14d"}.fa-compass:before{content:"\f14e"}.fa-toggle-down:before,.fa-caret-square-o-down:before{content:"\f150"}.fa-toggle-up:before,.fa-caret-square-o-up:before{content:"\f151"}.fa-toggle-right:before,.fa-caret-square-o-right:before{content:"\f152"}.fa-euro:before,.fa-eur:before{content:"\f153"}.fa-gbp:before{content:"\f154"}.fa-dollar:before,.fa-usd:before{content:"\f155"}.fa-rupee:before,.fa-inr:before{content:"\f156"}.fa-cny:before,.fa-rmb:before,.fa-yen:before,.fa-jpy:before{content:"\f157"}.fa-ruble:before,.fa-rouble:before,.fa-rub:before{content:"\f158"}.fa-won:before,.fa-krw:before{content:"\f159"}.fa-bitcoin:before,.fa-btc:before{content:"\f15a"}.fa-file:before{content:"\f15b"}.fa-file-text:before{content:"\f15c"}.fa-sort-alpha-asc:before{content:"\f15d"}.fa-sort-alpha-desc:before{content:"\f15e"}.fa-sort-amount-asc:before{content:"\f160"}.fa-sort-amount-desc:before{content:"\f161"}.fa-sort-numeric-asc:before{content:"\f162"}.fa-sort-numeric-desc:before{content:"\f163"}.fa-thumbs-up:before{content:"\f164"}.fa-thumbs-down:before{content:"\f165"}.fa-youtube-square:before{content:"\f166"}.fa-youtube:before{content:"\f167"}.fa-xing:before{content:"\f168"}.fa-xing-square:before{content:"\f169"}.fa-youtube-play:before{content:"\f16a"}.fa-dropbox:before{content:"\f16b"}.fa-stack-overflow:before{content:"\f16c"}.fa-instagram:before{content:"\f16d"}.fa-flickr:before{content:"\f16e"}.fa-adn:before{content:"\f170"}.fa-bitbucket:before{content:"\f171"}.fa-bitbucket-square:before{content:"\f172"}.fa-tumblr:before{content:"\f173"}.fa-tumblr-square:before{content:"\f174"}.fa-long-arrow-down:before{content:"\f175"}.fa-long-arrow-up:before{content:"\f176"}.fa-long-arrow-left:before{content:"\f177"}.fa-long-arrow-right:before{content:"\f178"}.fa-apple:before{content:"\f179"}.fa-windows:before{content:"\f17a"}.fa-android:before{content:"\f17b"}.fa-linux:before{content:"\f17c"}.fa-dribbble:before{content:"\f17d"}.fa-skype:before{content:"\f17e"}.fa-foursquare:before{content:"\f180"}.fa-trello:before{content:"\f181"}.fa-female:before{content:"\f182"}.fa-male:before{content:"\f183"}.fa-gittip:before{content:"\f184"}.fa-sun-o:before{content:"\f185"}.fa-moon-o:before{content:"\f186"}.fa-archive:before{content:"\f187"}.fa-bug:before{content:"\f188"}.fa-vk:before{content:"\f189"}.fa-weibo:before{content:"\f18a"}.fa-renren:before{content:"\f18b"}.fa-pagelines:before{content:"\f18c"}.fa-stack-exchange:before{content:"\f18d"}.fa-arrow-circle-o-right:before{content:"\f18e"}.fa-arrow-circle-o-left:before{content:"\f190"}.fa-toggle-left:before,.fa-caret-square-o-left:before{content:"\f191"}.fa-dot-circle-o:before{content:"\f192"}.fa-wheelchair:before{content:"\f193"}.fa-vimeo-square:before{content:"\f194"}.fa-turkish-lira:before,.fa-try:before{content:"\f195"}.fa-plus-square-o:before{content:"\f196"} diff --git a/docs/css/ie/PIE.htc b/docs/css/ie/PIE.htc deleted file mode 100644 index ca3b54700..000000000 --- a/docs/css/ie/PIE.htc +++ /dev/null @@ -1,96 +0,0 @@ - - - - - - - - - diff --git a/docs/css/ie/html5shiv.js b/docs/css/ie/html5shiv.js deleted file mode 100644 index dcf351c86..000000000 --- a/docs/css/ie/html5shiv.js +++ /dev/null @@ -1,8 +0,0 @@ -/* - HTML5 Shiv v3.6.2 | @afarkas @jdalton @jon_neal @rem | MIT/GPL2 Licensed -*/ -(function(l,f){function m(){var a=e.elements;return"string"==typeof a?a.split(" "):a}function i(a){var b=n[a[o]];b||(b={},h++,a[o]=h,n[h]=b);return b}function p(a,b,c){b||(b=f);if(g)return b.createElement(a);c||(c=i(b));b=c.cache[a]?c.cache[a].cloneNode():r.test(a)?(c.cache[a]=c.createElem(a)).cloneNode():c.createElem(a);return b.canHaveChildren&&!s.test(a)?c.frag.appendChild(b):b}function t(a,b){if(!b.cache)b.cache={},b.createElem=a.createElement,b.createFrag=a.createDocumentFragment,b.frag=b.createFrag(); -a.createElement=function(c){return!e.shivMethods?b.createElem(c):p(c,a,b)};a.createDocumentFragment=Function("h,f","return function(){var n=f.cloneNode(),c=n.createElement;h.shivMethods&&("+m().join().replace(/\w+/g,function(a){b.createElem(a);b.frag.createElement(a);return'c("'+a+'")'})+");return n}")(e,b.frag)}function q(a){a||(a=f);var b=i(a);if(e.shivCSS&&!j&&!b.hasCSS){var c,d=a;c=d.createElement("p");d=d.getElementsByTagName("head")[0]||d.documentElement;c.innerHTML="x"; -c=d.insertBefore(c.lastChild,d.firstChild);b.hasCSS=!!c}g||t(a,b);return a}var k=l.html5||{},s=/^<|^(?:button|map|select|textarea|object|iframe|option|optgroup)$/i,r=/^(?:a|b|code|div|fieldset|h1|h2|h3|h4|h5|h6|i|label|li|ol|p|q|span|strong|style|table|tbody|td|th|tr|ul)$/i,j,o="_html5shiv",h=0,n={},g;(function(){try{var a=f.createElement("a");a.innerHTML="";j="hidden"in a;var b;if(!(b=1==a.childNodes.length)){f.createElement("a");var c=f.createDocumentFragment();b="undefined"==typeof c.cloneNode|| -"undefined"==typeof c.createDocumentFragment||"undefined"==typeof c.createElement}g=b}catch(d){g=j=!0}})();var e={elements:k.elements||"abbr article aside audio bdi canvas data datalist details figcaption figure footer header hgroup main mark meter nav output progress section summary time video",version:"3.6.2",shivCSS:!1!==k.shivCSS,supportsUnknownElements:g,shivMethods:!1!==k.shivMethods,type:"default",shivDocument:q,createElement:p,createDocumentFragment:function(a,b){a||(a=f);if(g)return a.createDocumentFragment(); -for(var b=b||i(a),c=b.frag.cloneNode(),d=0,e=m(),h=e.length;d * { - float: left; - } - - .row:after { - content: ''; - display: block; - clear: both; - height: 0; - } - - .row:first-child > * { - padding-top: 0 !important; - } - - /* Normal */ - - .row > * { - /* padding-left: (gutters) */ - padding-left: 40px; - } - - .row + .row > * { - /* padding: (gutters) 0 0 (gutters) */ - padding: 40px 0 0 40px; - } - - .row { - /* margin-left: -(gutters) */ - margin-left: -40px; - } - - /* Flush */ - - .row.flush > * { - padding-left: 0; - } - - .row + .row.flush > * { - padding: 0; - } - - .row.flush { - margin-left: 0; - } - - /* Quarter */ - - .row.quarter > * { - /* padding-left: (gutters * 0.25) */ - padding-left: 10px; - } - - .row + .row.quarter > * { - /* padding: (gutters * 0.25) 0 0 (gutters * 0.25) */ - padding: 10px 0 0 10px; - } - - .row.quarter { - /* margin-left: -(gutters * 0.25) */ - margin-left: -10px; - } - - /* Half */ - - .row.half > * { - /* padding-left: (gutters * 0.5) */ - padding-left: 20px; - } - - .row + .row.half > * { - /* padding: (gutters * 0.5) 0 0 (gutters * 0.5) */ - padding: 20px 0 0 20px; - } - - .row.half { - /* margin-left: -(gutters * 0.5) */ - margin-left: -20px; - } - - /* One and (a) Half */ - - .row.oneandhalf > * { - /* padding-left: (gutters * 1.5) */ - padding-left: 60px; - } - - .row + .row.oneandhalf > * { - /* padding: (gutters * 1.5) 0 0 (gutters * 1.5) */ - padding: 60px 0 0 60px; - } - - .row.oneandhalf { - /* margin-left: -(gutters * 1.5) */ - margin-left: -60px; - } - - /* Double */ - - .row.double > * { - /* padding-left: (gutters * 2) */ - padding-left: 80px; - } - - .row + .row.double > * { - /* padding: (gutters * 2) 0 0 (gutters * 2) */ - padding: 80px 0 0 80px; - } - - .row.double { - /* margin-left: -(gutters * 2) */ - margin-left: -80px; - } \ No newline at end of file diff --git a/docs/css/style-mobile.css b/docs/css/style-mobile.css deleted file mode 100644 index c48699fa1..000000000 --- a/docs/css/style-mobile.css +++ /dev/null @@ -1,43 +0,0 @@ -/* - Solarize by TEMPLATED - templated.co @templatedco - Released for free under the Creative Commons Attribution 3.0 license (templated.co/license) -*/ - -/* Basic */ - - header.major h2 { - font-size: 1.6em; - } - - header.major .byline { - font-size: 1.1em; - } - - .wrapper { - padding: 4em 0em 2em 0em; - } - - .wrapper.style2 { - padding-bottom: 5em; - } - - .wrapper.style4 p { - margin-bottom: 1em; - } - - .wrapper.style5 img { - width: 70%; - } - -/* Lists */ - - ul.icons li a { - font-size: 1.6em; - } - -/* Main */ - - #main #sidebar .major h2 { - font-size: 1.6em; - } \ No newline at end of file diff --git a/docs/css/style-narrow.css b/docs/css/style-narrow.css deleted file mode 100644 index d810e7f4b..000000000 --- a/docs/css/style-narrow.css +++ /dev/null @@ -1,7 +0,0 @@ -/* - Solarize by TEMPLATED - templated.co @templatedco - Released for free under the Creative Commons Attribution 3.0 license (templated.co/license) -*/ - -/* Basic */ \ No newline at end of file diff --git a/docs/css/style-narrower.css b/docs/css/style-narrower.css deleted file mode 100644 index a4b7867b5..000000000 --- a/docs/css/style-narrower.css +++ /dev/null @@ -1,221 +0,0 @@ -/* - Solarize by TEMPLATED - templated.co @templatedco - Released for free under the Creative Commons Attribution 3.0 license (templated.co/license) -*/ - -/* Basic */ - - #skel-layers-wrapper { - padding-top: 44px; - } - - p { - line-height: 1.8em; - } - - header.major h2 { - font-size: 1.8em; - } - - header.major .byline { - font-size: 1.1em; - } - -/* Button */ - - input[type="submit"], - input[type="reset"], - input[type="button"], - .button { - display: block; - width: 100%; - max-width: 350px; - margin: 0em auto; - font-size: 1em; - font-weight: 600; - } - - .wrapper.style2 { - background: #f2f2f2; - text-align: center; - } - - .wrapper.style2 .major { - text-align: center !important; - } - - .wrapper.style2 .major h2 { - font-size: 1.4em; - } - - .wrapper.style2 .major .byline { - font-size: 1em; - } - - .wrapper.style2 h3 { - font-size: 1.2em; - } - - .wrapper.style3 .container { - padding-left: 1em; - padding-right: 1em; - } - - .wrapper.style3 p { - font-size: 1em; - } - - .wrapper.style4 { - padding-bottom: 2em; - } - - .wrapper.style5 { - padding-bottom: 4em; - } - - .wrapper.style5 img { - width: 30%; - height: 30%; - } - -/* Header */ - - #header { - display: none; - } - -/* Layers */ - - #titleBar { - background: #000; - height: 44px; - line-height: 44px; - } - - #titleBar .title { - display: block; - position: relative; - letter-spacing: 2px; - line-height: 44px; - font-size: 1.3em; - font-weight: 700; - text-align: center; - text-transform: uppercase; - color: #fff; - z-index: 1; - } - - #titleBar .toggle { - height: 60px; - left: 0; - position: absolute; - top: 0; - width: 80px; - z-index: 2; - } - - #titleBar .toggle:before { - content: "\f0c9"; - -moz-osx-font-smoothing: grayscale; - -webkit-font-smoothing: antialiased; - font-family: FontAwesome; - font-style: normal; - font-weight: normal; - text-transform: none !important; - } - - #titleBar .toggle:before { - background: none; - color: white; - display: block; - height: 44px; - line-height: 44px; - text-align: center; - width: 44px; - } - - #titleBar .toggle:after { - background: url("images/toggle.svg") 0px 0px no-repeat; - content: ''; - display: block; - height: 13px; - left: 10px; - opacity: 0.75; - position: absolute; - top: 15px; - width: 20px; - } - - #navPanel { - background: #000; - } - - #navPanel .link { - border-top: solid 1px rgba(0, 0, 0, 0.15); - color: #888; - display: block; - height: 44px; - line-height: 44px; - padding: 0 1em 0 1em; - text-decoration: none; - } - - #navPanel .link:first-child { - border-top: 0; - } - - #navPanel .link.depth-0 { - color: #fff; - } - - #navPanel .link .indent-1 { - display: inline-block; - width: 1em; - } - - #navPanel .link .indent-2 { - display: inline-block; - width: 2em; - } - - #navPanel .link .indent-3 { - display: inline-block; - width: 3em; - } - - #navPanel .link .indent-4 { - display: inline-block; - width: 4em; - } - - #navPanel .link .indent-5 { - display: inline-block; - width: 5em; - } - -/* Banner */ - - #banner { - padding: 6em 0em 3em 0em; - text-align: center; - text-transform: uppercase; - color: #FFF; - } - - #banner .container { - padding: 0em 1em; - } - - #banner h2 { - padding: 0.50em 0.50em; - font-size: 1.8em; - } - - #banner span, #banner p { - font-size: 1.1em; - } - - #banner span { - padding: 1em 0em; - } diff --git a/docs/css/style-normal.css b/docs/css/style-normal.css deleted file mode 100644 index 7a5ec66be..000000000 --- a/docs/css/style-normal.css +++ /dev/null @@ -1,11 +0,0 @@ -/* - Solarize by TEMPLATED - templated.co @templatedco - Released for free under the Creative Commons Attribution 3.0 license (templated.co/license) -*/ - -/* Basic */ - - body, input, select, textarea { - font-size: 11pt; - } \ No newline at end of file diff --git a/docs/css/style-wide.css b/docs/css/style-wide.css deleted file mode 100644 index d4b2ca723..000000000 --- a/docs/css/style-wide.css +++ /dev/null @@ -1,12 +0,0 @@ -/* - Solarize by TEMPLATED - templated.co @templatedco - Released for free under the Creative Commons Attribution 3.0 license (templated.co/license) -*/ - -/* Basic */ - - body, input, select, textarea { - font-size: 12pt; - line-height: 1.8em; - } \ No newline at end of file diff --git a/docs/css/style.css b/docs/css/style.css deleted file mode 100644 index dfb5e3d68..000000000 --- a/docs/css/style.css +++ /dev/null @@ -1,815 +0,0 @@ -@import url("https://fonts.googleapis.com/css?family=Montserrat:300,400,600,700,800"); - -/* Basic */ - - body { - background: #222222; - } - - body.loading * { - -moz-transition: none !important; - -webkit-transition: none !important; - -o-transition: none !important; - -ms-transition: none !important; - transition: none !important; - -moz-animation: none !important; - -webkit-animation: none !important; - -o-animation: none !important; - -ms-animation: none !important; - animation: none !important; - } - - body, input, select, textarea { - color: #424242; - font-family: 'Montserrat', sans-serif; - font-size: 12pt; - font-weight: 300; - line-height: 1.65em; - } - - a { - color: black; - text-decoration: underline; - } - - strong, b { - font-weight: 700; - } - - em, i { - font-style: italic; - } - - p, ul, ol, dl, table, blockquote { - margin: .5em 0 .75em 0; - } - - p { - line-height: 1.8em; - } - - p.medium { - font-size: 1.4em; - } - - h1, h2, h3, h4, h5, h6 { - color: inherit; - - } - - h1 a, h2 a, h3 a, h4 a, h5 a, h6 a { - color: inherit; - text-decoration: none; - } - -h1 > a:hover::after, -h2 > a:hover::after, -h3 > a:hover::after, -h4 > a:hover::after { - content: '\00A7'; - padding-left: 0.5em; - color: gray; -} - - h2 { - font-size: 1.5em; - font-weight: 800; - } - - h3 { - font-size: 1.25em; - font-weight: 700; - } - - h4 { - font-size: 1em; - font-weight: 700; - } - - - sub { - font-size: 0.6em; - position: relative; - top: 0.5em; - } - - sup { - font-size: 0.8em; - position: relative; - top: -0.5em; - } - - hr { - border-top: solid 1px #EA7826; - border: 0; - margin-bottom: 1.5em; - } - - blockquote { - border-left: solid 0.5em #EA7826; - font-style: italic; - padding: 1em 0 1em 2em; - } - - section.special, article.special { - text-align: center; - } - - header.major { - padding-bottom: 3em; - text-align: center; - text-transform: uppercase; - } - - header.major h2 { - font-size: 2.6em; - font-weight: 700; - } - - header.major .byline { - display: block; - padding-top: 1em; - letter-spacing: 1px; - font-size: 1.4em; - } - - footer > :last-child { - margin-bottom: 0; - } - - footer.major { - padding-top: 3em; - } - -/* Form - - input[type="text"], - input[type="password"], - input[type="email"], - textarea { - -moz-appearance: none; - -webkit-appearance: none; - -o-appearance: none; - -ms-appearance: none; - appearance: none; - background: none; - border: solid 1px #EA7826; - color: inherit; - display: block; - outline: 0; - padding: 0.75em; - text-decoration: none; - width: 100%; - } - - input[type="text"]:focus, - input[type="password"]:focus, - input[type="email"]:focus, - textarea:focus { - border-color: #EA7826; - } - - input[type="text"], - input[type="password"], - input[type="email"] { - line-height: 1em; - } - - ::-webkit-input-placeholder { - color: inherit; - opacity: 0.5; - position: relative; - top: 3px; - } - - :-moz-placeholder { - color: inherit; - opacity: 0.5; - } - - ::-moz-placeholder { - color: inherit; - opacity: 0.5; - } - - :-ms-input-placeholder { - color: inherit; - opacity: 0.5; - } - - .formerize-placeholder { - color: rgba(85, 85, 85, 0.5) !important; - } - -/* Image */ - - .image { - border: 0; - position: relative; - } - - img.profile { - width: 100%; - height: 100%; - } - .image.fit { - display: block; - } - - .image.fit img { - display: block; - width: 100%; - } - - .image.feature { - display: block; - margin: 0 0 2em 0; - } - - .image.feature img { - display: block; - width: 100%; - } - -/* Icon */ - - .icon { - position: relative; - } - - .icon:before { - content: ""; - -moz-osx-font-smoothing: grayscale; - -webkit-font-smoothing: antialiased; - font-family: Montserrat; - font-style: normal; - font-weight: normal; - text-transform: none !important; - } - - .icon > .label { - display: none; - } - -/* Lists */ - - ol.default { - list-style: decimal; - padding-left: 1.25em; - } - - ol.default li { - padding-left: 0.25em; - } - - ul.default { - margin: 0; - padding: 0; - list-style: square; - margin-left: 70px; - } - - ul.default li { - padding: 0.40em 0em; - } - - - - ul.actions { - cursor: default; - } - - ul.actions:last-child { - margin-bottom: 0; - } - - ul.actions li { - display: inline-block; - padding: 0 0 0 1.5em; - } - - ul.actions li:first-child { - padding: 0; - } - - ul.actions.vertical li { - display: block; - padding: 1.5em 0 0 0; - } - - ul.actions.vertical li:first-child { - padding: 0; - } - - ul.special-icons { - margin: 0em; - padding: 0em; - } - - ul.special-icons > li { - position: relative; - padding: 0.50em 0em 0.50em 0em; - } - - ul.special-icons > li:before { - position: absolute; - left: 0; - top: 2em; - display: block; - background: none; - font-size: 2em; - border-radius: 5px; - } - - ul.special-icons > li:first-child { - border-top: none; - } - - ul.special-icons h3 { - margin-bottom: 0.80em; - line-height: 2em; - text-transform: uppercase; - font-weight: 700; - font-size: 1.2em; - } - - ul.special-icons span { - line-height: 190%; - } - - ul.special-icons .fa { - float: left; - /* display: inline-block; */ - padding-right: 1em; - font-size: 1.4em; - color: #EA7826; - } - - ul.special-icons p { - padding-left: 2.7em; - } - -/* Tables */ - - table { - width: 100%; - } - - table.default { - width: 100%; - } - - table.default tbody tr { - border-bottom: solid 1px #EA7826; - } - - table.default td { - padding: 0.5em 1em 0.5em 1em; - } - - table.default th { - font-weight: 700; - padding: 0.5em 1em 0.5em 1em; - text-align: left; - } - - table.default thead { - background: #555555; - color: #fff; - } - -/* Button */ - - input[type="submit"], - input[type="reset"], - input[type="button"], - .button { - -moz-appearance: none; - -webkit-appearance: none; - -o-appearance: none; - -ms-appearance: none; - appearance: none; - background: none; - border-radius: none; - border: 2px solid; - border-color: #EA7826; - color: #EA7826; - cursor: pointer; - display: inline-block; - padding: 0.90em 1.2em; - letter-spacing: 1px; - text-align: center; - text-decoration: none; - text-transform: uppercase; - font-size: 1.4em; - -moz-transition: all 0.35s ease-in-out; - -webkit-transition: all 0.35s ease-in-out; - -o-transition: all 0.35s ease-in-out; - -ms-transition: all 0.35s ease-in-out; - transition: all 0.35s ease-in-out; - } - - input[type="submit"]:hover, - input[type="reset"]:hover, - input[type="button"]:hover, - .button:hover { - background: #EA7826; - color: white; - } - - input[type="submit"].alt, - input[type="reset"].alt, - input[type="button"].alt, - .button.alt { - border-color: inherit; - color: inherit; - } - - input[type="submit"].alt:hover, - input[type="reset"].alt:hover, - input[type="button"].alt:hover, - .button.alt:hover { - background: white; - color: #EA7826; - } - - input[type="submit"].fit, - input[type="reset"].fit, - input[type="button"].fit, - .button.fit { - width: 100%; - } - - input[type="submit"].small, - input[type="reset"].small, - input[type="button"].small, - .button.small { - font-size: 0.8em; - } - -/* Wrapper */ - - .wrapper { - padding: 6em 0em 4em 0em; - } - - - .wrapper.style1 { - padding: 0em; - background: #1b1b1b url(../images/banner.png) no-repeat; - background-size: contain; - background-position: center; - } - - .wrapper.style2 { - background: #f2f2f2; - } - - .wrapper.style2 .major { - text-align: left !important; - } - - .wrapper.style2 .major h2 { - display: block; - margin-bottom: 0.70em; - letter-spacing: 1px; - line-height: 1.4em; - text-transform: uppercase; - font-size: 1.8em; - font-weight: 400; - } - - .wrapper.style2 .major .byline { - letter-spacing: normal; - line-height: 1.6em; - text-transform: capitalize; - font-size: 1.4em; - } - - .wrapper.style2 h3 { - display: block; - margin-bottom: 1em; - letter-spacing: 1px; - line-height: 1.4em; - text-transform: uppercase; - font-size: 1.6em; - font-weight: 400; - } - - .wrapper.style3 { - padding-bottom: 6em; - background: #282661; - text-align: center; - color: white; - } - - .wrapper.style3 .container { - padding-left: 6em; - padding-right: 6em; - } - - .wrapper.style3 p { - font-size: 1.6em; - } - - .wrapper.style4 { - background: white; - } - - .wrapper.style5 { - background: #EA7826; - text-align: center; - color: white; - } - - .wrapper.style5 .image { - display: block; - width: 60%; - margin: 0em auto 2em auto; - } - - .wrapper.style5 .image img { - border-radius: 50%; - border: 10px solid; - border-color: white; - } - -/* Header */ - - #header { - color: white; - height: 4em; - background: rgba(0, 0, 0, 0.5); - } - - #header .container { - position: relative; - } - - .homepage #logo { - display: none; - } - - .homepage #nav { - position: static; - right: none; - text-align: right; - } - - .homepage a { - color: #ffffff; - } - - #logo { - line-height: 2em; - letter-spacing: 2px; - text-transform: uppercase; - font-size: 2em; - font-weight: 400; - } - - #logo h1 { - display: inline-block; - margin: 0; - padding: 0; - } - - #logo a { - color: inherit; - } - - #nav { - position: absolute; - top: 0; - right: 0; - } - - .logo-img { - float: left; - position: relative; - margin: 20px 15px 15px 10px; -} - - #nav > ul { - margin: 0; - } - - #nav > ul > li { - border-radius: 4px; - display: inline-block; - margin-left: 0.5em; - padding: 0 0.5em; - text-transform: uppercase; - font-weight: 600; - } - - #nav > ul > li a { - color: inherit; - line-height: 4em; - letter-spacing: 2px; - text-decoration: none; - text-transform: uppercase; - font-weight: 600; - font-size: 1em; - } - - #nav > ul > li:first-child { - margin-left: 0; - } - - #nav > ul > li.active a { - color: white; - } - - #nav > ul > li > ul { - display: none; - } - -/* Dropotron */ - - .dropotron { - top: 2em; - background: rgba(40, 38, 97, .85); - border-radius: 4px; - color: inherit; - min-width: 12em; - padding: 1em 0; - color: white; - } - - .dropotron > li { - line-height: 2em; - padding: 0 1em; - } - - .dropotron > li > a { - color: inherit; - letter-spacing: 2px; - text-decoration: none; - text-transform: uppercase; - } - - .dropotron > li.active > a, .dropotron > li:hover > a { - color: inherit; - } - - .dropotron.level-0 { - border-radius: 0 0 4px 4px; - font-size: 1em; - font-weight: 600; - margin-left: -0.5em; - } - -/* Banner */ - - #banner { - padding: 6em 0em 3em 0em; - text-align: center; - text-transform: uppercase; - color: white; - } - - #banner .container { - padding: 0em 8em; - height: 300px; - position: relative; - top: -75px; - - } - - #banner h2 { - display: inline-block; - padding: 0.50em 0.30em; - font-size: 3em; - font-weight: 800; - } - - #banner span, #banner p { - display: block; - letter-spacing: 1px; - text-transform: uppercase; - font-size: 1.6em; - font-weight: 800; - } - - #banner span { - padding: 1em 0em; - } - - #banner .banner-logo { - -webkit-filter: drop-shadow(3px 3px #333 ); - filter: drop-shadow(3px 3px #333); - max-width: 100%; - } - - #banner h3 { - text-shadow: 3px 3px #333; - } - -/* Main */ - - #main { - padding: 4em 0; - } - - #main #content .major, #main #sidebar .major { - text-align: left; - } - - #main #sidebar section { - margin-top: 4em; - } - - #main #sidebar section:first-child { - margin-top: 0; - } - - #main #sidebar .major { - padding-bottom: 2em; - } - - #main #sidebar .major h2 { - font-size: 1.8em; - } - -/* Footer */ - - #footer { - padding: 6em 0 2em 0; - text-align: center; - color: black; - background-image: linear-gradient(goldenrod,orange,red,purple) - } - - #footer a { - color: #000000; - text-decoration: underline; - } - - - #footer hr { - border-bottom: 1px solid; - opacity: .05; - } - - - #footer .major .byline { - opacity: .4; - } - - #footer .copyright { - margin-top: 3em; - text-align: center; - } - -/* Extra */ - - #extra1 h2 { - display: block; - margin-bottom: 1em; - letter-spacing: 1px; - line-height: 1.4em; - text-transform: uppercase; - font-size: 1.8em; - font-weight: 400; - } - - #extra1 h3 { - display: block; - margin-bottom: 1em; - letter-spacing: 1px; - line-height: 1.4em; - text-transform: uppercase; - font-size: 1.6em; - font-weight: 400; - } - -/* Team */ - - #team h3 { - margin-bottom: 0.50em; - letter-spacing: 2px; - text-transform: uppercase; - font-weight: 700; - } - -/* Embedded videos */ -.embedded.video { - display: block; - margin: auto; - padding: 1rem; -} - - -.label { - display: inline; - padding: .2em .6em .3em; - font-size: 75%; - font-weight: 700; - line-height: 1; - color: #EA7826; - text-align: center; - white-space: nowrap; - vertical-align: baseline; - border-radius: .25em; -} \ No newline at end of file diff --git a/docs/images/acah2021-cubed.png b/docs/images/acah2021-cubed.png deleted file mode 100644 index 3d29dbe10..000000000 Binary files a/docs/images/acah2021-cubed.png and /dev/null differ diff --git a/docs/images/acah2021-wide.png b/docs/images/acah2021-wide.png deleted file mode 100644 index 77ed470fe..000000000 Binary files a/docs/images/acah2021-wide.png and /dev/null differ diff --git a/docs/images/acasia2021-banner.png b/docs/images/acasia2021-banner.png deleted file mode 100644 index 9a2d79518..000000000 Binary files a/docs/images/acasia2021-banner.png and /dev/null differ diff --git a/docs/images/acasia2021-cubed.png b/docs/images/acasia2021-cubed.png deleted file mode 100644 index a27bff7c8..000000000 Binary files a/docs/images/acasia2021-cubed.png and /dev/null differ diff --git a/docs/images/acasia2021-wide.png b/docs/images/acasia2021-wide.png deleted file mode 100644 index 052c497bd..000000000 Binary files a/docs/images/acasia2021-wide.png and /dev/null differ diff --git a/docs/images/apachecon-2021.svg b/docs/images/apachecon-2021.svg deleted file mode 100644 index 53a4f2068..000000000 --- a/docs/images/apachecon-2021.svg +++ /dev/null @@ -1,1268 +0,0 @@ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - image/svg+xml - - - - - - - - - - - - - - - - - - - - - - 2021 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - September 21-23www.apachecon.com - - - - - - - - - - - - - - - - - 2021 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - September 21-23www.apachecon.com - - - - - - - - - - - - 2021 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - September 21-23www.apachecon.com - - diff --git a/docs/images/apachecon-asia-2021.svg b/docs/images/apachecon-asia-2021.svg deleted file mode 100644 index 3bd8db4db..000000000 --- a/docs/images/apachecon-asia-2021.svg +++ /dev/null @@ -1,1368 +0,0 @@ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - image/svg+xml - - - - - - - - - - - - - - - - - - - - - - ASIA 2021 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - August 6-8www.apachecon.com - - - - - - - - - - - - - - - - - ASIA 2021 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - August 6-8www.apachecon.com - - - - - - - - - - - - ASIA 2021 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - August 6-8www.apachecon.com - - diff --git a/docs/images/apachecon.png b/docs/images/apachecon.png deleted file mode 100644 index 01f4a8457..000000000 Binary files a/docs/images/apachecon.png and /dev/null differ diff --git a/docs/images/banner.png b/docs/images/banner.png deleted file mode 100644 index 9c71fece8..000000000 Binary files a/docs/images/banner.png and /dev/null differ diff --git a/docs/images/iconmonstr-facebook-6-96.png b/docs/images/iconmonstr-facebook-6-96.png deleted file mode 100644 index 53428941a..000000000 Binary files a/docs/images/iconmonstr-facebook-6-96.png and /dev/null differ diff --git a/docs/images/iconmonstr-twitter-1-96.png b/docs/images/iconmonstr-twitter-1-96.png deleted file mode 100644 index a2b6e1181..000000000 Binary files a/docs/images/iconmonstr-twitter-1-96.png and /dev/null differ diff --git a/docs/images/irc-icon.png b/docs/images/irc-icon.png deleted file mode 100644 index 59e06ebd1..000000000 Binary files a/docs/images/irc-icon.png and /dev/null differ diff --git a/docs/images/slack-icon.png b/docs/images/slack-icon.png deleted file mode 100644 index acff6db12..000000000 Binary files a/docs/images/slack-icon.png and /dev/null differ diff --git a/docs/index.html b/docs/index.html deleted file mode 100644 index 301453b41..000000000 --- a/docs/index.html +++ /dev/null @@ -1,100 +0,0 @@ - - - - - - - - - ApacheCon @Home - - - - - - - - - - - - - - - - - - - -
- - - - -
- -
-
-

Coming to your living room this September!

- -

Join us from the comfort of your own home for three days of Apache project content across dozens of categories, from Big Data to Fintech to Search to Tomcat and more.

- -

The call for presentations is NOW OPEN!

- -

Registration will open for the event in April!

- -
-
- - - - - - - - - - - diff --git a/docs/index.xml b/docs/index.xml deleted file mode 100644 index 192c69c9b..000000000 --- a/docs/index.xml +++ /dev/null @@ -1,2315 +0,0 @@ - - - - ApacheCon at Home 2021 - https://elek.github.io/acah2021/ - Recent content on ApacheCon at Home 2021 - Hugo -- gohugo.io - en-us - Thu, 12 Sep 2019 17:30:00 +0000 - - Apache httpd and TLS/SSL certificates validation - https://elek.github.io/acah2021/sessions/1136.html - Thu, 12 Sep 2019 17:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1136.html - We will look to 2 different things here, validation of the server certificate andnvalidation of the client certificates. For the server certificate we will show Let&rsquo;s encrypt and mod_md and speak aboutnthe new ACMEv2 protocol and OCSP stapling. For the client certificates we look to OCSP and other validations. Demo and quick start example will provided during the talk. - - - - Building an open source analytics distribution on Kubernetes that meets the needs of a modern data scientist. - https://elek.github.io/acah2021/sessions/1146.html - Thu, 12 Sep 2019 17:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1146.html - Modern machine learning libraries such as Jupyter, Kubeflow, PyTorch on technologies that are managed outside of the linux foundation, often forcing users to pay for cloud or hosted offerings to manage integration and tool dependencys. This challenges the ability of the ASF to continue blazing a trail of open innovation in the data science and analytics frontiers which are transforming so many businesses today. New projects in the ASF: such as Airflow, PredictionIO, SystemML, and NiFi - combined with technologies such as Spark and HDFS can be used to build an end-to-end analytics solution that is flexible and modular enough to meet the needs of modern data scientists, at pedabyte scale. - - - - Building Zhaopin's enterprise event bus based on Apache Pulsar - https://elek.github.io/acah2021/sessions/1189.html - Thu, 12 Sep 2019 17:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1189.html - Zhaopin.com is a Chinese online recruitment services provider. As a bilingual job board, Zhaopin.com has one of the largest selections of real-time job vacancies in China. All recruitment advertisements are provided by prominent Chinese and foreign companies and enterprises throughout China. Zhaopin.com provides professional HR services to over 2.2 million clients and its average daily page views are over 68 million.nApache Pulsar was developed to address several shortcomings of existing messaging systems including a lot of enterprise features, message durability, and lower message latency. - - - - Cassandra Traffic Management at Instagram - https://elek.github.io/acah2021/sessions/1042.html - Thu, 12 Sep 2019 17:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1042.html - Cassandra has been deployed for many years at Instagram and is still growing fast. Over the years, we’ve constantly improved the design of our infrastructure: we’ve introduced Cassandra proxy nodes to decouple the processing and storage workloads, we&rsquo;ve developed Rocksandra, a new storage engine relying on RocksDB that reduced the GC pressure and improved the efficiency of our clusters. Lately, we&rsquo;ve added an intermediate layer in our Cassandra infrastructure, where we could add many traffic optimizations to further improve our Cassandra clusters' efficiency and reliability. - - - - Creating a Stream Data Pipeline on Google Cloud Platform using Apache Beam - https://elek.github.io/acah2021/sessions/1322.html - Thu, 12 Sep 2019 17:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1322.html - We built a scalable and flexible stream data pipeline for our microservices on Google Cloud Platform (GCP), using Cloud Pub/Sub, Google Cloud Storage, BigQuery, and Cloud Dataflow, using Apache Beam. The stream data pipeline is working on the production system for Mercari, one of the biggest C2C e-commerce services in Japan. The pipeline currently accepts logs from 5+ microservices, and the number will increase soon. Our microservice architecture is based on the following three concepts: - - - - Data Movement & Integration at PayPal & LinkedIn using Apache Gobblin - https://elek.github.io/acah2021/sessions/1149.html - Thu, 12 Sep 2019 17:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1149.html - Data replication at PayPal drives various different business use-cases from fraud detection, user behavioral analysis, credit checks to lot of other offline business decisions. During this talk, we will present how Apache Gobblin empowers data movement and integrations at PayPal in partnership with LinkedIn to showcase all the recent features as well as the planned roadmap for the platform. Apache Gobblin is a distributed data integration framework that simplifies common aspects of big data integration such as data ingestion, replication, organization and lifecycle management for both streaming and batch data ecosystems. - - - - Production-ready stream data pipeline in Merpay, Inc - https://elek.github.io/acah2021/sessions/1329.html - Thu, 12 Sep 2019 17:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1329.html - We’ve started to provide our stream based data pipeline by using Google Cloud Dataflow and Apache Beam since 2018 fall. It collects event logs from microservices running on GKE, then transforms and forwards the logs to GCS and BigQuery to use for analytics, etc. As you know, implementing and operating streaming jobs are challenging. We’re encountered various issues during that time. I’d like to share our knowledge on development and operation perspective. - - - - Recent advances in Natural Language Processing and Deep Learning - https://elek.github.io/acah2021/sessions/1119.html - Thu, 12 Sep 2019 17:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1119.html - In recent years, Machine Learning (ML) approaches to Natural Language Processing (NLP) drastically changed. From a feature-based modelling, where NLP tasks are tackled by using a set of manually engineered features, we moved to the so-called representation learning: text is naturally represented as a sequence of symbols (e.g., words, characters, etc.) and the role of learning how these symbols should be represented is entirely left to the ML method. Deep Learning dominates such paradigmatic shift. - - - - Tips and Tricks - Getting the most out of Apache Cordova - https://elek.github.io/acah2021/sessions/1207.html - Thu, 12 Sep 2019 17:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1207.html - Mobile development is hard, no matter what tools you use. With thenmultitude of devices out there how can you maximize the chance to delightnyour users? Jesse has been actively contributing to hybrid mobilendevelopment projects and writing hybrid mobile apps for 10 years and willnshare his bag of tricks. Learn best practices so you can optimize yournworkflow, and develop the best possible mobile applications. - - - - Widening the Circle: The open source journey from user to contributor - https://elek.github.io/acah2021/sessions/1116.html - Thu, 12 Sep 2019 17:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1116.html - Open source draws its strength from the communities that use and build it. It&rsquo;s their diversity of perspective, skills, and accountability that makes individual projects stronger and builds a richer and solid ecosystem. But while open source is used by the entire world, that broad user community is not yet reflected in the contributor base. In fact, diversity in open source is significantly worse than in proprietary software. While we claim that contribution is open to all, clearly not everybody feels empowered or welcome to contribute to open source projects. - - - - A Beginner's Guide to HTTPS and TLS - https://elek.github.io/acah2021/sessions/1091.html - Thu, 12 Sep 2019 16:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1091.html - This talk will introduce you to the fundamentals of securing communications with HTTPS. We will start by explaining the basics of X.509 server and client certificates, certification authorities, and using the OpenSSL toolkit. The TLS/SSL protocol will be introduced and how it is used together with HTTP in order to provide for data encryption, integrity, and authentication. We will talk through some important configuration details, standard use cases, common pitfalls, known SSL vulnerabilities, and issues when using HTTPS. - - - - Cassandra 4.0: Our Most Stable Major Release - https://elek.github.io/acah2021/sessions/1095.html - Thu, 12 Sep 2019 16:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1095.html - In mid-2018 the Cassandra community committed to making Cassandra 4.0 the most stable major release of Cassandra in the project’s history. In September, the community shifted focus from feature work and development to ensuring the quality of the release. To this end, we have adopted several new approaches to testing and validation including the replaying of production traffic, code audits, and property-based testing. This talk will explore the methodologies we’ve adopted and the results of their application as well as costs of this level of commitment to testing and its benefits. - - - - Creating Custom Streaming Events to Learn Apache Beam - https://elek.github.io/acah2021/sessions/1328.html - Thu, 12 Sep 2019 16:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1328.html - Learning a streaming framework like Apache Beam is exciting but with &lsquo;Hello World!&rsquo; examples aren&rsquo;t fun. There aren&rsquo;t a lot of free and interesting streaming data sources for beginners to play with, it&rsquo;s very easy to give up learning if it&rsquo;s boring. To keep myself learning something new, I find myself need incentives and accomplishments to continue. If you are like me need some motivation to keep learning, this talk will give you some inspiration. - - - - Customer segmentation and personalization in websites/PWAs using Apache Unomi - https://elek.github.io/acah2021/sessions/1210.html - Thu, 12 Sep 2019 16:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1210.html - In this session, you will learn all that’s new with Apache Unomi, the open source Customer Data platform (that graduated this year) based on the Apache Karaf runtime, and all that’s happened since the last ApacheCon. You will discover how to easily integrate it with an existing website or SPA/PWA using its built-in web tracker, how to build customer segments and how to use the API to personalize the experience for your users. - - - - Event Sourcing with Spring Boot and Apache Kafka - https://elek.github.io/acah2021/sessions/1194.html - Thu, 12 Sep 2019 16:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1194.html - In our global economy, businesses must be nimble and often have to adapt quickly. As a result, many businesses structure their teams in an Agile way to keep up with this demand. With Domain Driven Design, it’s possible to quickly modify applications to accommodate changing business needs and easily integrate with disparate third-party systems. But what if you were able to use historical data and analytics to enhance your applications capabilities? - - - - OpenWhisk and Kubernetes based Mobile Backend as a Service - https://elek.github.io/acah2021/sessions/1206.html - Thu, 12 Sep 2019 16:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1206.html - So you&rsquo;ve been hearing a lot of buzz about Serverless tech in conjunction with Containers and Mobile, but what exactly is the OpenWhisk in reference to Kubernetes? Come find out at this session. Serverless has become the new style of coding, and it might be perfect to offload your mobile apps, and container based systems without incurring unnecessary costs.nGet an intro into serverless/function-as-a-service/cloud native technologies in the Mobile-Backend-as-a-Service (MBaaS) context and learn why startups and enterprises are so excited about using it. - - - - Riding Camel in Huawei Cloud - https://elek.github.io/acah2021/sessions/1130.html - Thu, 12 Sep 2019 16:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1130.html - Apache Camel provides a powerful EIP (Enterprise Integration Patterns) DSL to integrate tons of different middlewares. Now more and more enterprise user moving their application into cloud, it could be a challenge for them to bridge the public cloud service and the inhouse build application from scratch. With the help of Apache Camel, we provide an integration services on Huawei Cloud to address this across cloud integration issues and extend integration ability by leveraging the community collaboration. - - - - Robust Elastic Scaling of Deep Learning on Kubernates - https://elek.github.io/acah2021/sessions/1088.html - Thu, 12 Sep 2019 16:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1088.html - With the increasing popularity of deep learning workloads in academia, government and industry, several organizations have developed cloud-based deep learning platforms whose goals are to enable organizations utilize expensive resources effectively, and to share said resources among multiple teams in a fair and effective manner. We present the elastic scaling in the context of such large scale deep learning platforms. The traditional deep learning workload is the parallel job where can not fully utilize the cluster gpu resources. - - - - Running an Apache Project: 10 Traps and How to Avoid Them - https://elek.github.io/acah2021/sessions/1067.html - Thu, 12 Sep 2019 16:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1067.html - When you are starting on your open source adventure, there are lot ofnthings to learn that have very little to do with coding and insteadnrelate to interacting with people. Apache is, at its best, a group ofnpeople who are trying to share their experience and teach new projectsnand contributors how to successfully manage open source projects.nHowever, like the blind people each describing a part of an elephant,neach mentor brings their personal experience to the table, and thusncan give good, yet conflicting advice to new projects. - - - - Samza Portable Runner for Beam - https://elek.github.io/acah2021/sessions/1321.html - Thu, 12 Sep 2019 16:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1321.html - In this talk we are going to cover how we have leveraged portability of Beam and make Stream Processing in Python possible on top of Apache Samza. We will first touch points on Apache Samza in general, and how it stands out as the stream processing engine at LinkedIn that scales to over a trillion messages processed per day, with strong state support and flexible deployment models. Next we introduce Samza Runner for Beam, particularly the portable runner. - - - - What's New and Coming to Apache Hivemall: Building More Flexible Machine Learning Solution for Apache Hive and Spark - https://elek.github.io/acah2021/sessions/1158.html - Thu, 12 Sep 2019 16:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1158.html - Apache Hivemall is a scalable machine learning library for Apache Hive, Spark, and Pig. Hivemall allows us to apply a wealth of machine learning techniques to massive data stored in distributed storage by just writing a series of SQL-like queries. It provides classification, regression, recommendation, anomaly detection, and topic modeling functionalities in a scalable manner, along with a variety of auxiliary functions for data preprocessing and feature engineering. This talk demonstrates the Hivemall library with a special emphasis on its new features merged after the first Apache Incubator release. - - - - Adding AI to customer segmentation using Apache Unomi and Apache PredictionIO - https://elek.github.io/acah2021/sessions/1086.html - Thu, 12 Sep 2019 15:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1086.html - Apache Unomi is a Customer Data Platform that is easy to extend as it uses the Apache Karaf runtime. In this session, you will discover how to integrate Unomi with Apache PredictionIO to be able to leverage artificial intelligence algorithms to recommend segments or products to users as they interact with websites. Of course, as all of this will be bleeding-edge stuff, expect plenty of last-minute drama, half-finished code and hopefully some fun demos. - - - - Apache Cassandra Sidecar, let’s make C* attractive and easy to operate - https://elek.github.io/acah2021/sessions/1054.html - Thu, 12 Sep 2019 15:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1054.html - Cloud database offerings have expanded over the past decade, encompassing everything from virtualized machines in the cloud to entirely serverless databases. With this pace of innovation in the cloud ecosystem, Cassandra stands in a unique position to serve its users with unique advantages over any other system. This also puts Cassandra in an interesting position to compete with the ongoing innovations in the cloud. With the internal architecture and storage mechanism aside, citizen developers in the community are looking for several other operability aspects/ ecosystem around the services for the long term investments and benefits from such service. - - - - Apache Training, How it came to be and what it is about - https://elek.github.io/acah2021/sessions/1043.html - Thu, 12 Sep 2019 15:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1043.html - There are lots of companies offering training for, and around, the Apache ecosystem, as well as many other topics that all create their own training material. Keeping this material up to date, especially in the fast-moving open-source world is not an easy task and takes a lot of time and effort. And all the time someone not too far away is probably working on a very similar slide-deck to explain what ZooKeeper does&hellip; It is this fundamental issue that the Apache Training project was created to address - centralising training resources and making them easier to access and (re-)use. - - - - Configuring Apache Camel for the Cloud - https://elek.github.io/acah2021/sessions/1111.html - Thu, 12 Sep 2019 15:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1111.html - Cloud … check. Container … check. Orchestration … check. Now that you’ve got the basics it’s time to starting thinking about how all occupants of your new cloud infrastructure are going to communicate. And no mater what cloud, container or orchestration tool you choose, Apache Camel has what you need to get your system configured and connected. This talk will cover Camel components for cloud friendly configuration, communication, packaging, and deployment. - - - - Everything about Apache Marvin Platform - https://elek.github.io/acah2021/sessions/1082.html - Thu, 12 Sep 2019 15:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1082.html - Apache Marvin is an open source project that focuses on empowering data science teams to deliver industrial-grade applications supported by a high-scale, low-latency, language agnostic, and standardized architecture platform created to simplify the process of exploration and modeling in Machine Learning projects. Marvin aims at abstracting the complexities in the creation process of scalable, highly available, interoperable and maintainable predictive software. In this presentation, the speaker is going to talk about the architecture behind the platform and show the cooler features of Apache Marvin. - - - - Reverse Proxy Magic with Apache httpd 2.4 - https://elek.github.io/acah2021/sessions/1073.html - Thu, 12 Sep 2019 15:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1073.html - One of the best not-so-well-known features of Apache httpd is its incredible capability as a Reverse Proxy server. Whether as a front end to application servers such as Tomcat and PMP-FPM, or as a caching server for static content, or as a robust load-balancer with dynamic configuration, or all, Apache httpd is not only feature rich but easily matches or surpasses the performance of other proxies. - - - - Serverless Event Streaming with Pulsar Function: Use Cases and Best Practices - https://elek.github.io/acah2021/sessions/1193.html - Thu, 12 Sep 2019 15:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1193.html - As data continues to evolve, there are more and more requirements for calculating data. From servers to VMS to containers to serverless, the computing framework is constantly evolving with the needs of users. How to provide users with a fast, easy-to-deploy computing framework has become a question for everyone to think about. In this talk, we will introduce how Pulsar provides a powerful computing power based on the serverless computing architecture. - - - - Serverless: Multi-tenant Rule Engine Service Powered by Apache Karaf - https://elek.github.io/acah2021/sessions/1209.html - Thu, 12 Sep 2019 15:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1209.html - The Netflix media pipeline processes thousands of new shows and movies every day so that you can watch them on any device anywhere. We use a forward chaining rule engine to coordinate all of this work in multiple workflows. Hosting these workflows in a reliable, scalable and cost effective manner is a huge challenge at our scale. In this talk, we will introduce the design of Netflix’s next generation rule engine framework. - - - - The Power of Plugins - Developing plugins for Apache Cordova - https://elek.github.io/acah2021/sessions/1205.html - Thu, 12 Sep 2019 15:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1205.html - Plugins are the powerhouse inside Apache Cordova applications, this isnwhere all native functionality is exposed, and your mobile app trulynbecomes hybrid. We will explore the full scope of plugins in ApachenCordova, not just how they work, but how they are created, installed,ndiscovered and distributed. We will deep dive into concrete examples tondemonstrate the power of plugins. - - - - Unifying Batch and Stream Data Processing with Apache Calcite and Apache Beam - https://elek.github.io/acah2021/sessions/1320.html - Thu, 12 Sep 2019 15:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1320.html - At LinkedIn, we have a large and ever-expanding corpus of big data processing code written using batch processing frameworks like Pig, Hive, and Spark. These data processing pipelines produce derived data artifacts like metrics, dimensions, and features and are generally run at daily or hourly end-to-end latencies. For a subset of these artifacts, there often arises a need to produce them at a faster cadence, say minutely or even continuously. At that point, the common practice used to be that the developer would re-write the same derivation logic using a stream processing framework (Apache Samza in case of LinkedIn). - - - - Writing the Hazelcast Jet Runner - https://elek.github.io/acah2021/sessions/1327.html - Thu, 12 Sep 2019 15:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1327.html - Hazelcast Jet is a distributed data processing engine that threats all data as a stream. Jet is built on top of Hazelcast IMDG and thanks to this the Jet cluster can also play the role of the data source and sink. If you use Jet this way, you can achieve perfect data locality and top-of- the-class throughputs. Jet uses cooperative multithreading (comparable to green threads), it&rsquo;s processors correspond to standalone single-threaded tasks that process streams of data. - - - - Agile Integration - Cloud Native Application Development - https://elek.github.io/acah2021/sessions/1055.html - Thu, 12 Sep 2019 14:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1055.html - Cloud native application development is more than just container orchestration, when it comes to designing proper agile software architecture, there are many aspect that need to be taken into account. From simple microservices runtime, orchestration of core business, interacting with legacy, connecting with external SaaS application. To a more reactive system with events driven backbone, and also avoid data silos and how to deal with routing, versioning deployment strategy. Putting everything into a big picture, guide you through what next generation of cloud native architecture should be like and how everything works together. - - - - Apache Beam + Kotlin = ❤️ - https://elek.github.io/acah2021/sessions/1319.html - Thu, 12 Sep 2019 14:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1319.html - Kotlin is an Open Sourced, statically typed language for JVM and is mostly being favoured by Android Developers due to the many myriad features which enable more concise and cleaner code than Java without sacrificing performance or safety. By using Kotlin as our language of choice to build an apache beam pipeline, our development team not only ran into fewer errors than what we would have if we used Java; but with the help of extensively useful high level APIs present in Kotlin, we were also able to cut down on the development time required to build, test and deploy a new feature. - - - - Apache Fineract CN Mobile 2.0 - Supporting Branchless banking and promoting financial inclusion in the fragile states - https://elek.github.io/acah2021/sessions/1204.html - Thu, 12 Sep 2019 14:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1204.html - In today&rsquo;s world, we don&rsquo;t have an offline first open source mobile banking solution. Nowadays It is too important to have an offline first application because in many rural areas, it is exceptionally hard to find an internet connection, sometimes taking days to get connected if a natural disaster happens Still, there are some areas where internet connection is still not even present and despite the many global efforts to bring connectivity forward, there are regions where there is still no timeline for connectivity. - - - - Apache httpd 2.4 Overview - https://elek.github.io/acah2021/sessions/1057.html - Thu, 12 Sep 2019 14:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1057.html - All the Ins and Outs related to what&rsquo;s old, and NEW, in Apache httpd 2.4. If you haven&rsquo;t use httpd in awhile, this is a great way to find out all the new features and enhancements. And if you are a current user, you find out about how to tune httpd for peak performance - - - - Apache Karaf on cloud, the kloud initiative - https://elek.github.io/acah2021/sessions/1208.html - Thu, 12 Sep 2019 14:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1208.html - Apache Karaf has been seen for a while as an applications server. If it&rsquo;s actually the case for the standard distribution, it&rsquo;s very easy to create a Karaf powered standard/immutable distribution embedding your applications. In combination with other Karaf subprojects (Decanter, Cave, &hellip;), it provides a perfect applications runtime for the cloud.nThis talk will show how to start and focus on business code and then generate different distribution packages, up to a docker image ready to run on cloud. - - - - From an idea to an Apache TLP - https://elek.github.io/acah2021/sessions/1133.html - Thu, 12 Sep 2019 14:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1133.html - About 3 years ago, I had an idea of using Open-Source software to create the next generation of industrial it solutions. At ApacheCon 2017 in Miami I introduced this idea to the public with my &lsquo;Building SCADA systems with Apache Software&rsquo;. 2019 the Apache PLC4X project is an Apache TLP. In this talk I will not talk about technical details of the project itself, but all the steps we took on this journey from a community-building point of view. - - - - Hadoop Submarine Ecosystem: Bringing Machine Learning and Big Data world (YARN and Kubernetes) together - https://elek.github.io/acah2021/sessions/1196.html - Thu, 12 Sep 2019 14:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1196.html - &lsquo;Data Scientists focuses on developing ML models with applications such as TensorFlow / MXNet / Caffe / XGBoost and do not deep dive into the complexities of computing and storage to run ML/DL jobs. Today’s most of ETL processed data are stored in HDFS and cloud, and leveraging this data to design strong ML models is a big challenge for a data scientist. In Big Data ecosystem, most of the ETL and batch jobs are running on Spark and Hive which process/ingest data to the same data stores. - - - - Interactive querying of streams using Apache Pulsar - https://elek.github.io/acah2021/sessions/1192.html - Thu, 12 Sep 2019 14:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1192.html - As applications become more reliant on real-time data, streaming/messaging platforms have become more and more popular and crucial to any data pipeline. Currently, many streaming/messaging platforms are only used to access the most recent events from streams of data, however, there is tremendous value that can be unlocked if the full history of streams can be queried in an interactive fashion. Pulsar SQL is a query layer built on top of Apache Pulsar (a next-gen messaging platform), that enables users to dynamically query all streams, old and new, stored inside of Pulsar. - - - - Pricing Lyft rides with Apache Beam - a case study in migrating from a worker-based workflow to streaming - https://elek.github.io/acah2021/sessions/1326.html - Thu, 12 Sep 2019 14:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1326.html - Ride-sharing is a two-sided marketplace; balancing supply and demand with price in real time is critical to maintaining an efficient system. Dynamic pricing creates fairness for drivers (by raising rates when there is a lot of demand) and maintains good experiences for passengers (by satisfying pick-up time SLAs). This complex system makes real-time decisions using various data sources; machine learning models; and a streaming infrastructure for low latency, reliability and scalability. - - - - Reduce your Storage Costs with Transient Replication and Cheap Quorums - https://elek.github.io/acah2021/sessions/1188.html - Thu, 12 Sep 2019 14:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1188.html - nIn eventually consistent systems, when a node failures or network partition occurs, we’re presented with a trade-off: to execute a request and sacrifice consistency or reject execution and sacrifice availability. In such system, quorums, overlapping node subsets guaranteeing at least one node to hold the most recent value, can be a good middle-ground. We can tolerate failures and loss of connectivity for some nodes while still serving latest results. Quorum-based replication schemes incur high storage costs: we have to store redundant values on several nodes to guarantee enough copies are going to be available in case of failure. - - - - A Guided Walkthrough of the Apache Beam Go SDK - https://elek.github.io/acah2021/sessions/1318.html - Thu, 12 Sep 2019 13:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1318.html - This session will walk through the internals of what the Go SDK does to execute a pipeline. Describing the structure of the SDK, where code is located, and the why of the structure, and how it does it with the particular features of the Go language, which notably has strict types but no generics. Particular effort will be spent on what the Go SDK does for ease of use, and to reduce DoFn execution overhead. - - - - Advanced Apache Cordova Development - https://elek.github.io/acah2021/sessions/1203.html - Thu, 12 Sep 2019 13:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1203.html - Apache Cordova is a platform for building mobile apps using common Web technologies (HTML, CSS and JavaScript). Apache Cordova offers a set of APIs that allow the mobile app developers to utilize mobile native functions such as (Audio, Camera, Contacts …etc) using JavaScript. This session discusses and demonstrates the advanced aspects of Apache Cordova. Advanced aspects include extending Apache Cordova framework by creating custom plugins in order to access more device hardware features. - - - - Apache Groovy Update and Roadmap - https://elek.github.io/acah2021/sessions/1033.html - Thu, 12 Sep 2019 13:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1033.html - This talk looks at the latest features in Groovy from 2.5 and 3.0. This includes new AST transforms, new macro features, the new Parrot parser and a myriad of other new miscellaneous features. The talk outlines a broad roadmap of how the new features are planned to be rolled out and the system requirements for each version. - - - - Beaming Deep Learning with Ludwig - https://elek.github.io/acah2021/sessions/1325.html - Thu, 12 Sep 2019 13:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1325.html - Ludwig is a code-free Deep Learning toolbox based on TensorFlow open-sourced by Uber AI Labs. Ludwig is unique in its ability to help make deep learning easier to understand for non-experts and enable faster model improvement iteration cycles for experienced machine learning developers and researchers alike. By using Ludwig, experts and researchers can simplify the prototyping process and streamline data processing so that they can focus on developing deep learning architectures rather than data wrangling. - - - - Conquering Networked Applications using Ballerina Programming Language - https://elek.github.io/acah2021/sessions/1103.html - Thu, 12 Sep 2019 13:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1103.html - Ballerina is the next generation programming language, which redefined what means to be &lsquo;general purpose&rsquo;. Historically, programming languages concentrated on single machine execution, in a controlled environment, and also for good reason, because any other external interactions were out of the scope of a programming language, at least in the bygone days that is. But now, the communication network is something that is always there, and software often doesn&rsquo;t work alone. - - - - Kafka, Cassandra and Kubernetes at Scale - Real-time Anomaly detection on 19 billion events a day - https://elek.github.io/acah2021/sessions/1187.html - Thu, 12 Sep 2019 13:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1187.html - Apache Kafka, Apache Cassandra and Kubernetes are open source big data technologies enabling applications and business operations to scale massively and rapidly. While Kafka and Cassandra underpins the data layer of the stack providing capability to stream, disseminate, store and retrieve data at very low latency, Kubernetes is a container orchestration technology that helps in automated application deployment and scaling of application clusters. In this presentation, we will reveal how we architected a massive scale deployment of a streaming data pipeline with Kafka and Cassandra to cater to an example Anomaly detection application running on a Kubernetes cluster and generating and processing massive amount of events. - - - - One SQL to Rule Them All – a Syntactically Idiomatic Approach to Management of Streams and Tables - https://elek.github.io/acah2021/sessions/1191.html - Thu, 12 Sep 2019 13:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1191.html - Apache Calcite is a data management framework that includes a SQL parser and query optimizer. It is used by many projects that implement SQL processing capabilities, including Apache Beam and Apache Flink. Over the last years, members of these three communities had many discussions about the semantics and syntax of &lsquo;Streaming SQL&rsquo;. End of last year, we decided to formalize and summarize our views and ideas in paper that we submitted to the Industrial Track of the SIGMOD 2019 conference. - - - - The Apache Way May Save Us All - https://elek.github.io/acah2021/sessions/1041.html - Thu, 12 Sep 2019 13:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1041.html - The commercial software industry is plagued by numerous significant problems: Cost and schedule overruns, poor software quality, unrealistic expectations, products that fail to meet customer needs, and more. These problems are historic, chronic, and pervasive; they’ve been with us for decades, and changes to methodology haven’t had much effect in remedying them. By contrast, the Apache Software Foundation has been around for 20 years and the Apache Way - the framework of values and governance for Apache Software Foundation projects - has been the guiding force for dozens of highly successful software projects that are used in countless environments everywhere, run much of the Internet, and are probably used by every technical person in one way or another every day. - - - - The ASF and OFBiz - How Community Makes It All Possible - https://elek.github.io/acah2021/sessions/1065.html - Thu, 12 Sep 2019 13:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1065.html - Going back well before the ASF was formed, community has played a vital role in the development of open source software. With the formation of the ASF, we see a great example of open source community being formalized and actively facilitated. Although it can seem counterintuitive, especially to the uninitiated, placing &lsquo;Community over code&rsquo; helps ensure the ongoing relevance of open source as a movement, inviting individuals and organizations alike to invest in &lsquo;free&rsquo; software. - - - - Using Apache MXNet to Train and Deploy your Deep Learning model - https://elek.github.io/acah2021/sessions/1094.html - Thu, 12 Sep 2019 13:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1094.html - AI is evolving rapidly, and much of the recent advancement is driven by Deep Learning, a machine learning technique inspired by the inner-working of the human brain. In this session, we will discuss what deep learning is, and the new capabilities it enables. We will dive into a few computer vision and natural language models that are demonstrating super-human performance, and to integrate these models into your existing Java system leveraging Apache MXNet - an open source deep learning framework – and it&rsquo;s Java API. - - - - Apache Kafka vs. Integration Middleware (MQ, ETL, ESB) - Friends, Enemies or Frenemies? - https://elek.github.io/acah2021/sessions/1124.html - Thu, 12 Sep 2019 10:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1124.html - MQ, ETL and ESB middleware are often used as integration backbone between legacy applications, modern microservices and cloud services. This introduces several challenges and complexities like point-to-point integration or non-scalable architectures. This session discusses how to build a completely event-driven streaming platform leveraging Apache Kafka’s open source messaging, integration and streaming components to leverage distributed processing, fault-tolerance, rolling upgrades and the ability to reprocess events. nLearn the differences between an event-driven streaming platform leveraging Apache Kafka and middleware like MQ, ETL and ESBs – including best practices and anti-patterns, but also how these concepts and tools complement each other in an enterprise architecture. - - - - Apache NetBeans clever APIs for cross platform development - https://elek.github.io/acah2021/sessions/1106.html - Thu, 12 Sep 2019 10:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1106.html - Hybrid applications can be a great solution for cross platform development, but JavaScript is not everyones cup of coffee. With a small set of clever APIs and SPIs, NetBeans has created a lean cross platform framework, that allows you to code in Java, and run the application on any mobile or desktop platform, as hybrid, or fully native application. In this sessions I&rsquo;d like to unveil this hidden gem (improperly named &lsquo;HTML/Java&rsquo;), show some real life applications and discuss how this improves client side software development dramatically. - - - - Apache Toree: A Jupyter Kernel for Scala / Apache Spark - https://elek.github.io/acah2021/sessions/1089.html - Thu, 12 Sep 2019 10:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1089.html - Many data scientists are already making heavy usage of the Jupyter ecosystem for analyzing data using interactive notebooks. Apache Toree (incubating) is a Jupyter kernel designed that enables data scientists and data engineers to easily connect and leverage Apache Spark and its powerful APIs from a standard Jupyter notebook to execute their analytics workloads. In this talk, we will go over what&rsquo;s new with the most recent Apache Toree release. We will cover available magics and visualizations extensions that can be integrated with Toree to enable better data exploration and data visualizations. - - - - Driving dynamic Beam pipelines - https://elek.github.io/acah2021/sessions/1317.html - Thu, 12 Sep 2019 10:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1317.html - Using Apache Beam to get data in your data lake? In a agile company you don’t want to re-compile your ingestion pipeline every time a sprint finished. In this talk we go over all mechanisms and building blocks you need to make dynamic pipelines really work. We’ll see why schemas are so important. How do we get these schemas in our pipelines and discuss methods to protect ourselves from data corruption and incompatible schema evolution. - - - - Enterprise Hybrid Mobile App Development - https://elek.github.io/acah2021/sessions/1202.html - Thu, 12 Sep 2019 10:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1202.html - This session shows how to develop and run native mobile apps without needing to know much about the native platforms, such as Android and iOS, by leveraging a hybrid development framework such as Apache Cordova, which is capable of transforming an ordinary web app into a mobile app. The presentation discusses a real-life enterprise-level Android and iOS mobile app developed under pressure for a wireless service provider. It covers some of the technical details of the app as well as how to perform device-level debugging of the Android app, using adb. - - - - How China Search Giant Baidu adopt apache way inside its company - https://elek.github.io/acah2021/sessions/1092.html - Thu, 12 Sep 2019 10:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1092.html - Baidu is one of the biggest Internet company, it was founded 20 years ago with 10,000 more engineers now.nit began to embraces open source recently years, including adopting InnerSource and contributing to open source community. InnerSource is the use of apache ways inside the company. n As the leader of this program, I will talk about how this happened, and what's the challenges we faced and overcome. we need to set policies, define processes and enable tools , but the most important part is to cultivate open source culture, educate engineers what is apache way and how to cooperate inside company just like in open source community. - - - - How to Raise an Erudite Chatbot - https://elek.github.io/acah2021/sessions/1014.html - Thu, 12 Sep 2019 10:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1014.html - Availability of content and training sets is a major bottleneck for a chatbot development today. Relying on Apache OpenNLP and its sub-project OpenNLP.chatbot, we introduce a number of tools and components to design a chatbot and its training set to be knowledgeable and intelligent. n In this talk we will analyze the reasons it is so hard to find a chatbot demo today for a nontrivial task or to observe an intelligent behavior of a chatbot. - - - - OFBiz in the fashion Industry, an omnichannel approach - https://elek.github.io/acah2021/sessions/1186.html - Thu, 12 Sep 2019 10:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1186.html - The fashion industry works to meet the demand for apparel and dictates thentrends for what should be worn.nThis presentation will describe why Apache OFBiz is a great platform fornbuilding an ERP for the fashion industry and how it can be enhanced tonsupport its specific requirements. The various implementation steps of anconcrete project will be reviewed:n- Development of a custom eCommerce/B2B Responsive and Seo compliantnapplication using the Color-Size matrix feature.n- Integration of third-party applications such as retail, legacy Erp,nsocial network, WMS, Business Intelligence, mailing system. - - - - Running C* on Kubernetes with CassKop, the C* operator for K8S developed by Orange - https://elek.github.io/acah2021/sessions/1101.html - Thu, 12 Sep 2019 10:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1101.html - This presentation will describe the initial experience building and using CassKop, an operator developed for running Cassandra on top of Kubernetes. CassKop works as a usual K8S controller (reconcile the real state with a desired state) and automates the Cassandra operations through JMX. All the operations are launched by calling standard K8S APIs (kubectl apply …) or by using a K8S plugin (kubectl casskop …). CassKop is developed in GO, based on CoreOS operator-sdk framework. - - - - Stream processing for the masses with Beam, Python and Flink - https://elek.github.io/acah2021/sessions/1324.html - Thu, 12 Sep 2019 10:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1324.html - Python is a widely used programming language that is characterized by a low barrier to entry. As many other companies in the industry, Yelp has used Python as the main programming language to implement back-end services. Unfortunately, when it comes to stream processing Python presents several challenges, including performance limitations, lack of proper multi-threading support and limited framework options. -When the use cases for more advanced stream processing started to arise, at Yelp we decided to leverage Flink and introduce a Scala/Java stack for our Data Pipeline. - - - - Streamlining Streaming System Management with Apache Heron - https://elek.github.io/acah2021/sessions/1190.html - Thu, 12 Sep 2019 10:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1190.html - Apache Heron (Incubating) is a real-time, distributed, fault-tolerant stream processing engine. In 2018 it was donated to the Apache Foundation and is used for processing data sets in a way that allows systems to react to data as it is created. The big data era is here and it has become more critical than ever that organizations are able to make decisions based on the data they own in an efficient and accurate way. - - - - Apache Arrow + Apache Beam: A vision for cross-language, columnar data pipelines - https://elek.github.io/acah2021/sessions/1323.html - Thu, 12 Sep 2019 09:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1323.html - Apache Arrow and Apache Beam deserve to be together. It is a stated goal of both projects to provide mechanisms for performing data processing tasks across many languages. To accomplish this, both projects define a common binary “model” and provide multiple language implementations of that model. -However, they are approaching the overarching problem of cross-language data processing from two very different angles: Arrow is primarily concerned with moving columnar data across language boundaries with minimal overhead and providing optimized computation primitives to operate on that data, while Beam seeks to enable users to write scalable data pipelines in their language of choice and execute them on any distributed data-processing system. - - - - Apache Camel K: connect your Knative serverless applications with everything else - https://elek.github.io/acah2021/sessions/1059.html - Thu, 12 Sep 2019 09:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1059.html - When you start developing serverless applications in the real world, sooner rather than later you will need to talk to external (legacy) systems. In this talk you will discover how to leverage Apache Camel K to connect your Knative serverless applications with everything Camel can connect.nApache Camel K allows running Camel routes as serverless applications directly on top of any Kubernetes cluster, leveraging Knative serverless capabilities such as auto-scaling, scaling to zero, event-based communication in order to connect serverless functions and microservice with external systems. - - - - Apache Cordova In Action - https://elek.github.io/acah2021/sessions/1044.html - Thu, 12 Sep 2019 09:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1044.html - Apache Cordova is a platform for building Hybrid mobile apps using common Web technologies (HTML, CSS and JavaScript). Apache Cordova offers a set of APIs that allow the mobile app developers to utilize mobile native functions such as (Audio, Camera, Contacts …etc) using JavaScript. Ionic is one of the best mobile web application frameworks, which allows the web developers to develop neat mobile web applications. This session discusses why there is a need for Hybrid mobile development, the current challenges of mobile development, and how using Apache Cordova can help in overcoming many of these technical challenges. - - - - Apache NetBeans - Shameless Marketing Tool - https://elek.github.io/acah2021/sessions/1123.html - Thu, 12 Sep 2019 09:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1123.html - NetBeans has completed its transition to Apache and is now a top level Apache project with a strong and dedicated community and millions of users worldwide. NetBeans always had great support for Apache Maven &amp; having an IDE of our own is a great chance to further promote Apache projects to a large audience. In this session I&rsquo;ll show you how to plugin your own language, tool, library, server, database or framework and make it easy for developers to get started with them. - - - - Apache OFBiz in the Cloud(s) - https://elek.github.io/acah2021/sessions/1140.html - Thu, 12 Sep 2019 09:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1140.html - Cloud computing greatly simplifies provisioning, scaling, reliability and maintenance: these aspects are particularly important for Enterprise applications like Apache OFBiz. This talk will provide effective tips, based on experience and lessons learned in the field, to enable OFBiz for the cloud. Topics covered include OFBiz specific solutions for:n- deployments in AWSn- images - AMI, Dockern- continuous deploymentsn- configurations for load balancersn- session management - sticky sessions, distributed sessions (with Redis)n- database options for the cloud - RDS (AWS), Google Cloud Spanner - - - - Continuous Machine and Deep Learning at Scale with Apache Ignite - https://elek.github.io/acah2021/sessions/1125.html - Thu, 12 Sep 2019 09:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1125.html - With most machine learning (ML) and deep learning (DL) frameworks, it can take hours to move data, and hours to train models. It&rsquo;s also hard to scale, with data sets increasingly being larger than the capacity of any single server. The size of the data also makes it hard to incrementally test and retrain models in near real-time to improve results. Learn how Apache Ignite and GridGain help to address these limitations with model training and execution, and help achieve near-real-time, continuous learning. - - - - How Netflix debugs and fixes Apache Cassandra when it breaks - https://elek.github.io/acah2021/sessions/1006.html - Thu, 12 Sep 2019 09:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1006.html - Netflix relies on Apache Cassandra as a critical source of truth database, and while Cassandra is a remarkably resilient database, it does, ever so occasionally, break. This talk explores how complex Cassandra deployments fail in production, but more importantly, the techniques, tools, and approaches our distributed systems engineers use to debug and mitigate these failures. We will first cover software-based failure modes that come either from our software or the software that Cassandra builds upon. - - - - How to Slide Your Release Pass the Incubator - https://elek.github.io/acah2021/sessions/1096.html - Thu, 12 Sep 2019 09:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1096.html - All podling releases need to be voted on by the incubator PMC before being released to the world. I&rsquo;ll go through what the incubator PMC looks for in every release and what you can do to make it pass that IPMC vote and get your project one step closer to graduation. More importantly I&rsquo;ll cover where you can get help if you need it. In this talk, I&rsquo;ll describe current incubator and ASF policy, recent changes that you may not be aware of, and go into detail the legal requirements of common open source licenses and the best way to assemble your NOTICE and LICENSE files. - - - - Running Apache Flink and Apache Beam on Kubernetes - https://elek.github.io/acah2021/sessions/1316.html - Thu, 12 Sep 2019 09:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1316.html - Access to real-time data is increasingly important for many organizations. At Lyft, we process millions of events per second in real-time to compute prices, balance marketplace dynamics, detect fraud, among many other use cases. To do so, we run dozens of Apache Flink and Apache Beam pipelines. Flink provides a powerful framework that makes it easy for non-experts to write correct, high-scale streaming jobs, while Beam extends that power to our large base of Python programmers. - - - - Spoilt for Choice – Kafka Streams vs. KSQL for Stream Processing on top of Apache Kafka - https://elek.github.io/acah2021/sessions/1195.html - Thu, 12 Sep 2019 09:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1195.html - Apache Kafka is a de facto standard streaming data processing platform. It is widely deployed as event streaming platform. Part of Kafka is its stream processing API “Kafka Streams”. In addition, the Kafka ecosystem now offers KSQL, a declarative, SQL-like stream processing language that lets you define powerful stream-processing applications easily. What once took some moderately sophisticated Java code can now be done at the command line with a familiar and eminently approachable syntax. - - - - TVM: An Automated End-to-End Optimizing Compiler for Deep Learning - https://elek.github.io/acah2021/sessions/1121.html - Thu, 12 Sep 2019 09:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1121.html - There is an increasing need to bring machine learning to a wide diversity of hardware devices from the datacenter to the edge. Current frameworks rely on vendor-specific operator libraries and optimize for a narrow range of server-class GPUs. Deploying workloads to new platforms such as mobile phones, embedded devices, and accelerators (e.g., FPGAs, ASICs) requires significant manual effort. In this talk, we will talk about Apache TVM &ndash; an end to end optimizing deep compiler stack that brings deep learning models on diverse hardware back-ends that are competitive with state-of-the-art hand-tuned libraries for low-power CPU, mobile GPU, and server-class GPUs. - - - - Apache Camel K: a cloud-native integration platform - https://elek.github.io/acah2021/sessions/1142.html - Wed, 11 Sep 2019 17:45:00 +0000 - - https://elek.github.io/acah2021/sessions/1142.html - In this session we are going to introduce the latest innovation from the Apache Camel community: Camel K, a lightweight integration platform, born on Kubernetes, with serverless superpowers.nCamel K enables developers that want to integrate systems to directly write Camel DSL code in the cloud, with a great developer experience and really fast turnaround times.nYou’ll see Camel K in action with a live coding demo that will explore the main features that it provides. - - - - Apache httpd reverse proxy and Tomcat: - https://elek.github.io/acah2021/sessions/1156.html - Wed, 11 Sep 2019 17:45:00 +0000 - - https://elek.github.io/acah2021/sessions/1156.html - Although mostly known as a fast and reliable web server, Apache httpd also excels as a reverse proxy.nIn this session find out how to setup httpd as a reverse proxy, how to connect to Tomcat using HTTP and AJP.nWe will also look to the full feature list of Apache httpd proxying capability. - - - - Building Data Platform for your Next Meetup Event with Apache Foundation on Cloud - https://elek.github.io/acah2021/sessions/1138.html - Wed, 11 Sep 2019 17:45:00 +0000 - - https://elek.github.io/acah2021/sessions/1138.html - One of the challenges at Meetup is how to build a scalable, reliable and efficient data platform to help our ML team builds models that recommend events fit your interests. With emerging sophisticated batch and streaming frameworks and cloud solutions, our data platform went through massive changes in the past two years. In this talk, I&rsquo;ll discuss the evolution of how Meetup data platform utilizes Apache-based data systems, including Sqoop, Hive, Flume, Spark, Flink, Beam, Airflow. - - - - Consistent Cassandra schema changes in Elassandra - https://elek.github.io/acah2021/sessions/1070.html - Wed, 11 Sep 2019 17:45:00 +0000 - - https://elek.github.io/acah2021/sessions/1070.html - As described in CASSANDRA-10699 (Make schema alterations strongly consistent), concurrent schema changes can still lead to schema disagreement in Cassandra 3.0. In order to properly support Elasticsearch dynamic mapping in Elassandra, we will see how multiple schema changes are validated on a working copy of the Cassandra schema, and applied in an atomic update to all nodes if a light weight transaction succeed, thus avoiding concurrent schema changes issue. I will also explain how we have taken advantage of Cassandra table extensions to store the Elasticsearch mapping directly into the CQL schema with several benefits. - - - - Exploring github data using Apache Drill running on ARM64 - https://elek.github.io/acah2021/sessions/1155.html - Wed, 11 Sep 2019 17:45:00 +0000 - - https://elek.github.io/acah2021/sessions/1155.html - Inspired by Google&rsquo;s Felipe Hoffa&rsquo;s effort/talk, this attempt is to replicate an usecase of exploring github data using Apache Drill and lessons learnt. Apache Drill is a distributed MPP query layer that supports SQL and alternative query languages against NoSQL and Hadoop data storage systems. The talk will highlight steps to build Apache Drill from source code, and showcase a demo analyzing terabytes of github data (issues and comments), running on a cluster of 3 node ARM64 servers. - - - - Happiness is a hybrid cloud with Apache Cassandra™ - https://elek.github.io/acah2021/sessions/1028.html - Wed, 11 Sep 2019 17:45:00 +0000 - - https://elek.github.io/acah2021/sessions/1028.html - The world&rsquo;s data and the speed with which we expect to access it continues to grow rapidly. Apache Cassandra was built for hybrid cloud implementations and that makes it easy for companies to fully exploit their cloud strategies without having to re-architect their applications or change their systems for data management. Sounds too good to be true? A lot of big companies have been doing it for years and you can too! - - - - Podling's Shark Tank - https://elek.github.io/acah2021/sessions/1038.html - Wed, 11 Sep 2019 17:45:00 +0000 - - https://elek.github.io/acah2021/sessions/1038.html - Is it a panel? Is it a talk? It is a Podling&rsquo;s Shark Tank! Back by popular demand with even sharkier judges! What is it, you ask? Well, this is just like Shark Tank TV show (think speed dating between entrepreneurs and investors) but instead of “Squirrel Boss” and “Man Candle” you&rsquo;ll be hearing pitches for Apache Incubator projects. Also instead of Mark Cuban and Kevin O&rsquo;Leary you&rsquo;ll be pitching to the panel of ASF old timers (trying to convince them that your project is worthy of their esteemed attention and endorsement). - - - - Simple, Portable data pipelines with Apache Beam SQL - https://elek.github.io/acah2021/sessions/1309.html - Wed, 11 Sep 2019 17:45:00 +0000 - - https://elek.github.io/acah2021/sessions/1309.html - Apache Beam is a unified data processing framework, allowing you to write batch and streaming pipelines that run anywhere, including Apache Flink, Apache Spark, and Google Cloud Dataflow. With the SQL extension you can now write a pipeline in pure SQL. If you need more, you can write user defined functions in Java or even embed SQL into your existing Java pipeline. This talk will start with a demo pipeline written in pure SQL. - - - - Web-based Interactive Big Data Visualization - https://elek.github.io/acah2021/sessions/1183.html - Wed, 11 Sep 2019 17:45:00 +0000 - - https://elek.github.io/acah2021/sessions/1183.html - Data visualization is a challenging job to do in the Web-based environment due to the limited rendering efficiency of Web browsers. Rendering big data and providing user interaction in real time is even more challenging. nIn this speech, we are going to introduce how we made this possible with ECharts, a popular Web-based data visualization tool. Generally speaking, the solution lies in the following four aspects:n1. Simplifying shapes and rendering faster but with a similar result. - - - - Apache Kibble Crash Course - https://elek.github.io/acah2021/sessions/1099.html - Wed, 11 Sep 2019 16:45:00 +0000 - - https://elek.github.io/acah2021/sessions/1099.html - This presentation aims at showcasing the Apache Kibble project, getting new users set up with their own Kibble instance and getting started with project and community metrics. During this talk, the audience will learn what Apache Kibble is, what it does, HOW it does it, and how you can easily get set up in a few minutes and start reviewing metrics from your own project. We will also show a demo instance with a select handful of Apache projects, and dive into both single community metrics as well as comparative measuring across foundations. - - - - Combining schema-on-read and schema-provisioning in Apache Drill - https://elek.github.io/acah2021/sessions/1185.html - Wed, 11 Sep 2019 16:45:00 +0000 - - https://elek.github.io/acah2021/sessions/1185.html - The data generated from IoT devices, machine logs and similar sources is often semi-structured or unstructured. This poses a challenge for traditional schema-on-write systems that require a fixed schema up-front for querying. Modern analytic applications often have ad hoc usage patterns and require tremendous flexibility over how this data can be consumed. Further, they demand that the data be made available for querying soon after it has landed in their data platform – which may be a distributed file system or NoSQL database or something similar. - - - - Developing new IO connectors in Apache Beam - https://elek.github.io/acah2021/sessions/1308.html - Wed, 11 Sep 2019 16:45:00 +0000 - - https://elek.github.io/acah2021/sessions/1308.html - One of the key components in any data processing system is IO connectors. These fundamental blocks allow to read and write data, which is stored in different type of sources, in a unified and distributed way. In this sense, Apache Beam is not an exception - it provides a rich API to develop a new connector with your favorite SDK and easily integrate it with Beam runners. In this talk we are going to show you how to write your own IO connector (in Java). - - - - Efficient data integrations with the Medley platform - https://elek.github.io/acah2021/sessions/1034.html - Wed, 11 Sep 2019 16:45:00 +0000 - - https://elek.github.io/acah2021/sessions/1034.html - In this talk, we will introduce the Medley data integration platform used at Uber. Medley is based on the Apache Camel framework, and centered around integrating internal systems efficiently and with great precision on multiple network environments. We will focus on the concept of reusability of integrations, where integrations can be securely shared by implementing integrations as web services. Also, implementing business entity change detection with publish/subscribe to changes to allow event based integrations based on data state changes. - - - - From Postgres to an In-Memory Grid with Apache Ignite - https://elek.github.io/acah2021/sessions/1147.html - Wed, 11 Sep 2019 16:45:00 +0000 - - https://elek.github.io/acah2021/sessions/1147.html - This talk describes how a food tech company with 50k merchants was at a critical situation by delivering events to devices by using events polling on a Postgres database, and how it could solve this problem by moving the events repository to Apache Ignite using distributed, in-memory SQL. The change also allowed the company to be on the right track to support more than 500k simultaneous merchants. - - - - I Love Lucee - https://elek.github.io/acah2021/sessions/1148.html - Wed, 11 Sep 2019 16:45:00 +0000 - - https://elek.github.io/acah2021/sessions/1148.html - Lucee is a free and open source application server that implements the CFML language, which was popularized by the commercial product ColdFusion. It allows for rapid application development, and is implemented as a JSP Servlet so it is deployed inside a Servlet Container like Apache Tomcat. The Lucee website: https://lucee.org/ GitHub Repo:nhttps://github.com/lucee/Lucee Support Forum: https://dev.lucee.org/ - - - - Instaclustr's Open Source Tools For Cassandra - LDAP/Kerberos, Prometheus Exporter, Debug Tooling and K8s Operator - https://elek.github.io/acah2021/sessions/1026.html - Wed, 11 Sep 2019 16:45:00 +0000 - - https://elek.github.io/acah2021/sessions/1026.html - This session walks devs through Instaclustr&rsquo;s Cassandra tools and how they add key functionality and ease-of-use to their deployments.n- An LDAP authenticator plug-in for CassandranThe open source LDAP authenticator plug-in works closely with the existing CassandraAuthorizer implementation. The plug-in enables developers to quickly reap the benefits of secure LDAP authentication without the need to write their own solutions, and to transition to using the authenticator with zero downtime. - A Kerberos authenticator plug-in for CassandranThe open source Kerberos authenticator plug-in enables Cassandra users to leverage Kerberos’ industry-leading secure authentication and true single sign-on capabilities. - - - - Mission Critical Cassandra in Financial Services, The When/What/How Experiences for ALL! - https://elek.github.io/acah2021/sessions/1110.html - Wed, 11 Sep 2019 16:45:00 +0000 - - https://elek.github.io/acah2021/sessions/1110.html - Apache Cassandra is a database product that is most notably known for being highly scalable and fault tolerant. Since these characteristics are something that is desirable in almost every use case, it attracts attention from developers up through executives when faced with decisions on how to architect mission critical systems. If these traits are so desirable, and Apache Cassandra makes these capabilities available to end users, then shouldn’t we use Cassandra for all our data storage needs? - - - - Samza 1.0: How we scaled stream processing at LinkedIn - https://elek.github.io/acah2021/sessions/1182.html - Wed, 11 Sep 2019 16:45:00 +0000 - - https://elek.github.io/acah2021/sessions/1182.html - &lsquo;At LinkedIn, we generate a staggering 2 trillion events to Kafka each day. nThe ingested events are processed by our streaming platform powered by Apache Samza. nWe currently have over 3000 applications in production that leverage Samza at LinkedIn. nThe use-cases include the LinkedIn feed, anomaly detection, combating fraud, profile notifications, realtime analytics and many more. In this talk, we will describe our journey in building and scaling a distributed stream nprocessing system over these years. - - - - (Apache) Drill-ing into Collections of HDF5 Files - https://elek.github.io/acah2021/sessions/1069.html - Wed, 11 Sep 2019 15:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1069.html - This talk is about building bridges between two ecosystems, Apache and HDF5.nHDF5 is a widely used storage container for complex data from embedded devicesnto supercomputers, and is developed and maintained as FOSS by The HDF Group.nWhile it is easy to imagine the potential benefits of making HDF5 containersnaccessible from the various Apache frameworks, there are several technicalnchallenges to overcome, and what makes a &lsquo;good&rsquo; integration is by no meansnobvious. It requires the input and collaboration of experts from both sides. - - - - Apache Tomcat / HTTPD Configuration Management - https://elek.github.io/acah2021/sessions/1122.html - Wed, 11 Sep 2019 15:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1122.html - Does your organization have tons of applications and containers? Has this fleet grown to be completely unmanageable? Do undocumented changes cause downtime and pain for your operations? Using modern configuration management tools and practices (Chef and CI/CD), Mastercard created a set of cookbooks that builds, deploys and manages model driven middleware infrastructure. Join us for a talk where we will discuss what is and why you should do configuration management, how we chose to model the middleware configurations, examples of how to test the heck out of the whole assembly and how this methodology brings everything a step closer to “gitops”. - - - - Building S3 over Ozone : Making a Cloud Native File System - https://elek.github.io/acah2021/sessions/1104.html - Wed, 11 Sep 2019 15:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1104.html - The AWS S3 protocol is the defacto interface for modern object stores. Ozone supports S3 protocol as the first-class notion in Ozone. For all practical purposes, a user of S3 can start using Ozone without any change to code or tools. S3 protocol support offered by Ozone is strongly consistent, so users don’t need to run sidekick tools like S3Guard when running big data applications like Apache Spark, Apache YARN or Apache Hive. - - - - DataSketches - The Required Toolkit for the Analysis of Big Data - https://elek.github.io/acah2021/sessions/1181.html - Wed, 11 Sep 2019 15:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1181.html - In​ the analysis of b​ig data there are often problem queries that don’t scale because they require huge compute resources to generate exact results, or don’t parallelize well. Examples include c​ount-distinct, ​quantiles, most frequent items, joins, matrix computations, and graph analysis. Algorithms that can produce accuracy guaranteed approximate answers for these problem queries are a required toolkit for modern analysis systems that need to process massive amounts of data​ quickly. For interactive queries there may not be other viable alternatives, and in the case of real­-time streams, these specialized algorithms, called stochastic, s​treaming, sublinear algorithms,​ or &lsquo;s​ketches&rsquo;,​ are the only known solution. - - - - Day to day with Cassandra: The weirdest and complex situations we found! - https://elek.github.io/acah2021/sessions/1126.html - Wed, 11 Sep 2019 15:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1126.html - Every Cassandra operator has been hit with a couple of weird/complex cases that don&rsquo;t fit the normal expected failure situation. It can be a problem in hardware, software, networking, operator mistake, or a mix of it all. In this talk we will go through a compilation of such cases that we faced. How do they appear, how did we debug them and how did we fix them. We expect this to be a walk through weird, fun cases and sharing knowledge on the situations and on the fixing of such problems - - - - Lessons from the Apache Beam Journey - https://elek.github.io/acah2021/sessions/1024.html - Wed, 11 Sep 2019 15:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1024.html - Apache Beam is a big data processing framework that can run on multiple big data engines, including Apache Spark, Apache Flink, Apache Samza, Apache Apex, and a number of non-Apache OSS engines and proprietary engines. After situating Beam in the OSS ecosystem, I will share lessons from Apache Beam&rsquo;s journey, including topics such as: - how do you onboard newcomers to the Apache Way?n - how is open source different / the same as internal corporate work? - - - - Mythbusting with Apache Cassandra - https://elek.github.io/acah2021/sessions/1302.html - Wed, 11 Sep 2019 15:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1302.html - As a mature operational database with a 10+ year history as an Apache project, Cassandra has a long history with many twists and turns. In this talk, we’ll look at common misgivings and misunderstandings that we encounter in conversations with developers and get to the bottom of the story: - Cassandra is a “columnar” (or “column-oriented”) database. Porting my relational data model to Cassandra will work well. Cassandra is a niche database that is only good for narrow set of use cases. - - - - State of integration with Apache Camel - https://elek.github.io/acah2021/sessions/1141.html - Wed, 11 Sep 2019 15:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1141.html - Apache Camel is the leading open source integration framework, which has been around for over a decade. In this talk we will look back in history, to understand how the integration landscape has evolved from EAI, SOA, and ESB architectures up to microservices, and now with modern serverless and cloud native platforms. Apache Camel has been along for the ride. And we will look to the future and see how the latest release v3 of Apache Camel, is aimed for running modern cloud native workloads with Camel K. - - - - The journey of building a Beam runner based on Spark structured streaming framework - https://elek.github.io/acah2021/sessions/1307.html - Wed, 11 Sep 2019 15:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1307.html - Apache Beam provides a unified programming model to execute batch and streaming pipelines on all the popular big data engines. The translation layer from Beam to the chosen big data engine is called a runner. The current runner for Apache Spark is based on the RDD/DStream framework. However, there is an ongoing work to move it to Spark next generation framework a.k.a structured streaming. This talk will present why structured streaming is a good fit for Apache Beam, why it is worth the effort, and will give some feedback on how Apache Beam has solved the challenge, what the tough points and the sweet points were. - - - - Cassandra at Instagram 2019 - https://elek.github.io/acah2021/sessions/1074.html - Wed, 11 Sep 2019 14:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1074.html - At Instagram, our mission is to capture and share the world&rsquo;s moments. Our app is used by over 1 billion people monthly; this creates a lot of challenging data needs. We use Cassandra behind a lot of different product features. In this presentation, I will talk about how we use Cassandra to serve our critical use cases; the improvements we made in last several years to make sure Cassandra can meet our low latency, high scalability requirements; and future plans. - - - - Declarative Benchmarking of Cassandra and Its Data Models - https://elek.github.io/acah2021/sessions/1025.html - Wed, 11 Sep 2019 14:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1025.html - You have made changes to Cassandra code base. How do you benchmark these changes for scalability and correctness, including different data models (schema), easily? You have created a Cassandra schema for your service. How do you ensure this is scalable? How can you emulate application specific CQL queries, with specified distribution, to validate scale of your schema and associated data scalability without having to code your whole application? I am the author of the NDBench CQL Plugin tool, which was built at Netflix to address these needs and more, Declaratively. - - - - Drilling Security Data - https://elek.github.io/acah2021/sessions/1050.html - Wed, 11 Sep 2019 14:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1050.html - Security data is often challenging to analyze because it comes in a variety of formats that are difficult and time consuming to parse, requiring a myriad of various tools to analyze this data. Additionally, bringing this data together to correlate multiple data sets can be difficult and extremely time consuming. But what if all your data could be queried with single tool, using a common language? This talk will demonstrate how to use Apache Drill&rsquo;s enormous analytic power on security data sets and visualize this data using Apache Superset. - - - - HTTP/2, HTTP/3 and SSL/TLS State of Art in Our Servers - https://elek.github.io/acah2021/sessions/1135.html - Wed, 11 Sep 2019 14:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1135.html - A new protocol is getting ready HTTP/3 we will look to where we are with it in our serves.nThe &lsquo;old&rsquo; HTTP/2 protocol and the corresponding TLS/SSL are common to Traffic Server, HTTP Server and Tomcat.nThe presentation will shortly explain the new protocol and the ALPN extensions and look to the state of the those in our 3 servers and show the common parts and the specifics of each servers.nA demo configuration of each server will be run. - - - - Protect your Private Data in your Hadoop Clusters with ORC Column Encryption - https://elek.github.io/acah2021/sessions/1066.html - Wed, 11 Sep 2019 14:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1066.html - Fine-grained data protection at a column level in data lake environments has become a mandatory requirement to demonstrate compliance with multiple local and international regulations across many industries today. ORC is a self-describing type-aware columnar file format designed for Hadoop workloads that provides optimized streaming reads, but with integrated support for finding required rows quickly. In this talk, we will outline the progress made in Apache community for adding fine-grained column level encryption natively into ORC format that will also provide capabilities to mask or redact data on write while protecting sensitive column metadata such as statistics to avoid information leakage. - - - - SPARQL at scale with Apache Rya - https://elek.github.io/acah2021/sessions/1080.html - Wed, 11 Sep 2019 14:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1080.html - The Resource Description Framework (RDF) is a standard model for storing graph data. While the standard was initially created for storing meta-data about the World Wide Web, its flexible format made it a popular choice for storing many different types of information. With the explosive increase in the size of available data, scalable solutions are needed to efficiently store and query very large RDF graphs within big data architectures. Apache Rya (incubating) is a scalable database management system designed for storing and searching very large RDF data. - - - - The Rise of Open Source Software (and Apache projects) in China - https://elek.github.io/acah2021/sessions/1021.html - Wed, 11 Sep 2019 14:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1021.html - This talk will cover the evolution, the current status and the outlook of the China open source landscape and ecosystem. This talk will also share the summary and analysis of varies open source software survey reports for China, the largest software developers market in the world. The Apache Software Foundation&rsquo;s projects and their communities play a critical role in leading and developing the OSS ecosystems in China where some leading IT enterprises and individuals, such as Huawei, Alibaba, Baidu, Tencent, Kyligence, SkyWalking, etc. - - - - Using Relational Cache to Boost Apache Spark SQL - https://elek.github.io/acah2021/sessions/1180.html - Wed, 11 Sep 2019 14:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1180.html - Apache Spark has become a popular engine for data analysis in industry, providing a handful SQL interface and processing data from various data sources. While many of our customers are using Apache Spark as an interactive query platform to meet their business requirement, one of the most common complaints we heard is that spark is not quite &lsquo;interactive&rsquo; while data goes big. In traditional database systems, we can utilize materialized views to accelerate query processing, through pre-computation and query plan rewriting. - - - - Using the Timer and State API to solve times series use case in Apache Beam - https://elek.github.io/acah2021/sessions/1306.html - Wed, 11 Sep 2019 14:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1306.html - Time series processing requires ordered processing of data points, often across windows boundaries. Tricky stuff for distributed processing systems, in this chat we look at how to solve for these problems using a combination of Global Windows, State and the Timer API' with Apache Beam. - - - - Apache When You’re Not In Charge - https://elek.github.io/acah2021/sessions/1151.html - Wed, 11 Sep 2019 12:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1151.html - Apache has a 20 year history of building resilient and sustainable communities of software developers through the Apache Way. While some companies are adopting Inner sourcing, other companies still have a taboo surrounding open source culture. Many of us work at these open source agnostic companies and don’t have manager in our title. Are those of us in those companies doomed to a life of Dilbert? You don’t have to be! - - - - Friendlier Communities Through Infrastructure: Apache Beam’s Journey - https://elek.github.io/acah2021/sessions/1305.html - Wed, 11 Sep 2019 12:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1305.html - In this talk we will discuss the infrastructure Apache Beam has built to create a positive contributor experience. Over the last few years we have developed tooling, refactored code, instrumented Jenkins builds, collected metrics, created a metrics dashboard, and involved the community in order to ensure happy contributors. We will discuss the importance of engineering productivity and tooling in developing strong communities, where we have succeeded, and where we still have work to do. - - - - Game Changing Operational Visibility tools in Apache Cassandra 4.0 - https://elek.github.io/acah2021/sessions/1184.html - Wed, 11 Sep 2019 12:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1184.html - In this talk, we will review the new and upcomming tools for Apache Cassandra operations. This talk will go over some of the new and soon to be tools for operational visibility. Specific areas covered will include:n- Different operational tasks will be exposed with an HTTP interface with the sidecarn- Virtual tables provides visibility to the state of a node with cqlshn- New commands added to nodetool and JMX See how these improvements and additions will give operation teams faster diagnostics to assist in discovering problematic clients and data models. - - - - Lessons Learned from Leveraging Real-Time Power Consumption Data with Apache Kudu - https://elek.github.io/acah2021/sessions/1179.html - Wed, 11 Sep 2019 12:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1179.html - IoT is widely spreading in various industries, and various sensor devices generate large amounts of data in real-time. Such big data is used for visualization, machine learning, and other data analysis. Kudu is one of the data stores for such use cases which require fast inserts and efficient scans. Apache HBase is also often used to handle time series data such as sensor data. We also evaluated HBase for smart meter data a few years ago. - - - - Modernizing the Nordstrom Supply Chain: Efficient Order Fulfillment using JanusGraph, Cassandra and Spark - https://elek.github.io/acah2021/sessions/1102.html - Wed, 11 Sep 2019 12:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1102.html - DESCRIPTION: Today&rsquo;s Nordstrom customers expect order fulfillment options shipping to the most convenient locations in the timeframes that best fit their schedules. This session offers an inside look at how Nordstrom&rsquo;s Supply Chain Technology team combined an extensive network of brick and mortar facilities with a custom software stack built atop JanusGraph to present fulfillment options tailored to each individual order. ABSTRACT: Over the last decade, the volume of online retail sales as a proportion of overall retail sales has nearly tripled, and there are no signs of this trend slowing down. - - - - Packaging Tomcat for Linux Distributions - https://elek.github.io/acah2021/sessions/1004.html - Wed, 11 Sep 2019 12:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1004.html - What is the difference between the ASF Tomcat distribution and the various flavors of tomcat packages on Linux? Most users don’t know! In this discussion Coty will explain how the tomcat packages distributed by a couple flavors of Linux (Red Hat and Debian/Ubuntu) are structured and why the container is provided that way. Coty also hopes to get feedback from the audience and to discuss how these Linux distributions can better provide tomcat to users. - - - - Schema-Controlled HDFS Column Encryption and Use Cases - https://elek.github.io/acah2021/sessions/1132.html - Wed, 11 Sep 2019 12:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1132.html - Motivation: Access control via encryption improves security coverage compared with traditional enforcement in the access path because encryption can prevent invalid accesses from any angle. Finer-grained access control at the column-level is needed because in a typical big dataset, only a few columns are sensitive and need to be protected, and different columns could have different sensitivities and a different set of eligible readers. Design: With encryption features in columnar file format like Apache Parquet, column access control via encryption becomes possible. - - - - Supporting Cassandra In-House – Our Story !! - https://elek.github.io/acah2021/sessions/1008.html - Wed, 11 Sep 2019 12:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1008.html - In this presentation, senior Cassandra Architects at Ericsson will share their vast experience in supporting around 100 Cassandra deployments in production. They will share the key challenges and best practices with respect to Cassandra operations, maintenance and support. There is plenty to learn when the team talks about many problems which they faced in production and how they fixed each one of them successfully by providing interesting solutions. Audience: All Cassandra users especially Cassandra operators and administrators. - - - - What’s Surprising about Apache Drill and Why That’s a Challenge - https://elek.github.io/acah2021/sessions/1052.html - Wed, 11 Sep 2019 12:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1052.html - &lsquo;Apache Drill has some very surprising characteristics and, more importantly, it enables Drill users to do some surprising things. It’s no longer surprising to be able to do standard SQL in a highly distributed and large scale system - there is an entire class of modern tools that do this including Apache Hive, Presto or Spark SQL. But Drill has other capabilities that are surprising and make it stand apart from its class. - - - - Apache Hadoop 3.x State of The Union and Upgrade Guidance - https://elek.github.io/acah2021/sessions/1049.html - Wed, 11 Sep 2019 11:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1049.html - &lsquo;Apache Hadoop YARN is the modern Distributed Operating System for big data applications. It morphed the Hadoop compute layer to be a common resource-management platform that can host a wide variety of applications. Many organizations leverage YARN in building their applications on top of Hadoop without themselves repeatedly worrying about resource management, isolation, multi-tenancy issues etc. The Hadoop Distributed File System (HDFS) is the primary data storage system used by Hadoop applications. - - - - Apache Tomcat, your webapp, and the Graal - https://elek.github.io/acah2021/sessions/1048.html - Wed, 11 Sep 2019 11:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1048.html - Graal&rsquo;s native image generation has been the focus of a lot of attention recently, despite being in the very early stages. This session will look at the Tomcat embedded packaging, how to handle your webapp, and will demo use of Graal and Tomcat in a container environment. The session will mostly be tutorial/demo oriented, after a brief overview of the concept and tools used. - - - - Building BigData Query Optimization with Apache Calcite – Best Practices from Alibaba MaxCompute - https://elek.github.io/acah2021/sessions/1178.html - Wed, 11 Sep 2019 11:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1178.html - MaxCompute is a large scale, distributed big data platform of Alibaba, which provides Exabyte storage capacity and massive computing power through tens of thousands of commodity machines. The system supports SQL-like declarative language for advanced query and analysis on web-scale data set. Millions of jobs with hundreds of petabytes of data are processed every day, powering mission critical business within Alibaba, including e-Commerce, mobile payment, logistics, etc. Query optimizer plays a key part in determining the optimal execution plan. - - - - Managing Trillions of Rows with Aplomb (well, actually with Drill) - https://elek.github.io/acah2021/sessions/1022.html - Wed, 11 Sep 2019 11:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1022.html - &lsquo;Ingesting lots of data isn’t very hard any more. Ingesting it on a critical schedule, within strict time bounds while minimizing the risk of bogus data showing up is much harder. In practice, grownup data ingestion and access requires the following capabilities * Incoming data can be fully ingested into our working dataset but hidden from users until all quality checks are completedn * Individual batches of data can be released atomicallyn * Any indexing updates should also appear appear atomicallyn * Expiring data should disappear atomically either according to ingest batch or precise time bounds Apache Drill provides several capabilities that make it much easier to meet these goals. - - - - Portable Spark Runner: Running Beam Pipelines Written in Python and Go with Spark - https://elek.github.io/acah2021/sessions/1304.html - Wed, 11 Sep 2019 11:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1304.html - Apache Spark is the most popular open source analytics engine for large-scale data processing. Spark is not only a mature system, but thanks to its support of multiple resource managers like Hadoop, Mesos, and Kubernetes it has become a popular choice for both batch and streaming workloads in the industry. Apache Beam has included a Spark runner since its inception to allow users to execute Beam pipelines on Spark, but until recently the Spark runner could only execute pipelines written in Java. - - - - The Apache TinkerPop4 Virtual Machine - https://elek.github.io/acah2021/sessions/1060.html - Wed, 11 Sep 2019 11:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1060.html - The Apache TinkerPop team is working on the next major version of the project &ndash; TinkerPop4. TinkerPop4 advances the technology developed in TinkerPop3 with a new virtual machine architecture that enables any query language to control any processing engine manipulating any data structure. A query language can support a compiler to TinkerPop4&rsquo;s Turing Complete bytecode specification. Compiled bytecode is then translated into an intermediate functional representation that an integrated processing engine will ultimately transform into a processor-specific execution plan. - - - - The State of Diversity in Open Source & the ASF - https://elek.github.io/acah2021/sessions/1023.html - Wed, 11 Sep 2019 11:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1023.html - I&rsquo;d like to cover the landscape of D&amp;I in tech, then in open source, and then in the ASF. I&rsquo;d like to discuss the dimensions used to evaluate our current status and I&rsquo;d like to discuss the efforts we&rsquo;re working on to changer things. I will like to give this talk in the general track to a) celebrate the 20th Anniversary by reflecting back in our wins on diversity and inclusion and b) make it part of the foundation&rsquo;s big picture - - - - Two years in the making: What’s new with Apache Cassandra 4.0? - https://elek.github.io/acah2021/sessions/1027.html - Wed, 11 Sep 2019 11:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1027.html - At almost two years in the making Apache Cassandra 4.0 is almost here. With a focus on performance and stability, it is full of interesting features. This talk takes you through a tour of the new features and performance improvements. From Zero Copy Faster Streaming support to Virtual Tables and Audit Logging, learn how Apache Cassandra 4.0 has departed from previous major releases to be focused on the features required by the most demanding users. - - - - Using the TLP toolchain as a crystal ball for your cluster - https://elek.github.io/acah2021/sessions/1068.html - Wed, 11 Sep 2019 11:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1068.html - Cassandra cluster management is hard. Understanding how your Cassandra data model will hold up in production over a period of time can be tricky. If that is not tough enough, understanding how a change to a Cassandra setting will affect your cluster can be be difficult. Knowing how your data model or setting change will perform under a production data load can prevent performance degradation or worse nodes going down. In the last year, The Last Pickle has invested the time to develop the tooling necessary to create test clusters in AWS, as well as a scalable stress tool which can run pre-configured workloads. - - - - Anomaly Prediction in Apache CloudStack - https://elek.github.io/acah2021/sessions/1017.html - Tue, 10 Sep 2019 18:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1017.html - There is a chance of system, storage and network failures in the CloudStack environment at any point of time, which is unknown to the end users and administrators. The administrators are highly depend on the alert and monitoring system to raise an alarm on failures, upon the action is being taken. This can down the operations and will impact the end user services. The prediction of such failures ahead in CloudStack environment would be required instead of depending on the monitoring systems. - - - - Apache@Google, an account of the success of open source practices at Google Cloud - https://elek.github.io/acah2021/sessions/1333.html - Tue, 10 Sep 2019 18:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1333.html - We&rsquo;ll share with the Apache Community the journey that Google Cloud has had since we donated the first project to the foundation. We&rsquo;ll share our experiences from the commercial vendor point of view and also what we have learned while trying to embrace an &ldquo;open source first&rdquo; culture. - - - - Building and running cloud-native Cassandra - https://elek.github.io/acah2021/sessions/1075.html - Tue, 10 Sep 2019 18:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1075.html - This session will re-evaluate Cassandra’s relationship with runtime and build systems, pointing out ways that the existing systems fall down, and identifying avenues for improvement. Over the past few years, a number of platforms have emerged for running user code. Container runtimes like Docker, container orchestrators such as Kubernetes, and metrics collections agents like Prometheus and Spectator have all gained popularity and mind-share. Cassandra functionality such as metrics, bootstrapping, and monitoring integrates with the newer paradigms, but in an ad-hoc and improvised fashion. - - - - Hadoop Storage in the Cloud Native Era - https://elek.github.io/acah2021/sessions/1177.html - Tue, 10 Sep 2019 18:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1177.html - Hadoop was born much earlier than the Cloud Native era. But the question is still the same: what can it offer in the time of Kubernetes, containerization and hybrid clouds? Apache Hadoop Ozone is a new subproject of Hadoop. It is based a generic low-level binary layer, the Hadoop Distributed Data Storage (HDDS) and a S3 compatible Object Store implementation on top of it. The separation of the data storage layer has multiple benefits: it can help to solve the “small files problem” and increase the scalability of the Hadoop storage clusters but it also helps us to provide access to the same data via multiple interfaces. - - - - How to Grow Communities, Not Just Code - https://elek.github.io/acah2021/sessions/1084.html - Tue, 10 Sep 2019 18:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1084.html - Good code isn&rsquo;t enough for a successful open source project. First of all, only you know how to use what you&rsquo;ve made. Maybe it&rsquo;s time for a little UI and UX help? At the very least some documentation! Next, how is anyone else going to find what you&rsquo;ve created? And that&rsquo;s only the beginning. Ruth Suehle, manager of Red Hat&rsquo;s Open Source and Standards community leadership team, will take you through examples of the best and the worst, from projects large and small, to help you see what you need beyond your code to build a successful open source project and community. - - - - Inside Apache Druid: Built for High-Performance Real-Time Analytics - https://elek.github.io/acah2021/sessions/1171.html - Tue, 10 Sep 2019 18:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1171.html - Interactive applications are replacing traditional reporting interfaces as the preferred means for organizations to derive value from their datasets. An interactive user experience requires latency on the order of milliseconds. Cluster computing frameworks such as Apache Hadoop or Apache Spark are tremendously beneficial in processing and deriving insights from data. However, high query latency makes these frameworks sub-optimal for interactive applications. Alternatively, the use of relational databases and key/value stores as dedicated query layers can reduce latency, but these approaches suffer many drawbacks for analytic use cases. - - - - Introducing Secure Boot in Apache CloudStack - https://elek.github.io/acah2021/sessions/1036.html - Tue, 10 Sep 2019 18:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1036.html - nExisting CloudStack supports legacy boot loader BIOS by default for Guest VM’s booting on Hypervisor(XEN, VMware, KVM etc). In Virtualization, hypervisor emulates the hardware for the Guest VM. BIOS {Basic Input-Output system} is a low-level software. The BIOS loads when a VM starts up, and the BIOS is responsible for waking up your VM’s hardware(emulated) components, ensures they are functioning properly, and then runs the boot loader that boots Windows or whatever other operating system have installed. - - - - Standardizing Industrial IoT with Apache PLC4X, Apache Daffodil, and DFDL - https://elek.github.io/acah2021/sessions/1081.html - Tue, 10 Sep 2019 18:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1081.html - Industrial IoT is made up of many different programmable logic controllers (PLCs) that communicate with often incompatible and proprietary data formats. Apache PLC4X standardizes access to these PLCs by creating a set of libraries across a variety of languages to read and write PLC data with a unified API. Unfortunately, developing and maintaining support for these different data formats, especially across multiple languages, can be burdensome. In this talk we discuss how we can describe PLC data using the open standard Data Format Description Language (DFDL), how Apache Daffodil uses these descriptions to transform data to a standard and easily consumable representation, and an approach to extend PLC4X to use Daffodil and DFDL descriptions to reduce the effort needed to support a variety of PLCs and languages. - - - - Testing Contributions at Scale - https://elek.github.io/acah2021/sessions/1211.html - Tue, 10 Sep 2019 18:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1211.html - Time may be one of the most valuable resources in a project. Automating code reviews to allow for other tasks is a crucial goal for many communities. This talk will cover one way many Apache projects have significantly increased code quality and contribution feedback while simultaneously doing more with less. - - - - Tomcat from a cluster to a cloud. - https://elek.github.io/acah2021/sessions/1083.html - Tue, 10 Sep 2019 18:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1083.html - Using Tomcat in a cluster and in a cloud. nWe start by looking how to configure tomcat to get a cluster andnthen explore the problems and solutions to have distributed applications running in a cloud. Most cloud providers now have a Kubernetes API.nWe will look to what we have to add to Tomcat to have a decent cloud supportnfor monitoring, tracing and operating on the cloud.nWe will show how to use all the pieces. - - - - Watching TV on the command line with Apache Traffic Control - https://elek.github.io/acah2021/sessions/1005.html - Tue, 10 Sep 2019 18:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1005.html - Watch closely as a CDN built with the components of Apache Traffic Control comes alive in the delivery of fragmented ABR video. This session will walk through every step in a CDN transaction using command line tools that simulate what real-world clients do behind-the-scenes. - - - - Apache Doris (incubating) -- A simple and single tightly coupled olap system - https://elek.github.io/acah2021/sessions/1170.html - Tue, 10 Sep 2019 17:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1170.html - Apache Doris (incubating) is an MPP-based interactive SQL data warehousing for reporting and analysis. It is open-sourced by Baidu. Doris mainly integrates the technology of Google Mesa and Apache Impala. Unlike other popular SQL-on-Hadoop systems, Doris is designed to be a simple and single tightly coupled system, not depending on other systems. Doris not only provides high concurrent low latency point query performance, but also provides high throughput queries of ad-hoc analysis. - - - - Apache Training - Contributing more than just code - https://elek.github.io/acah2021/sessions/1172.html - Tue, 10 Sep 2019 17:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1172.html - Does it seem strange to you that we collectively collaborate on code but training material is produced individually in private? Why would each company or person produce their own material when it can be sourced from a central location, under a business friendly license, and built on and modified? Or perhaps you just see better ways of producing content, then come along and listen to what the Apache Training project is doing. - - - - CloudStack Monitoring - In-Depth Checks For Production Deployments - https://elek.github.io/acah2021/sessions/3002.html - Tue, 10 Sep 2019 17:00:00 +0000 - - https://elek.github.io/acah2021/sessions/3002.html - This talk is for CloudStack administrators who&rsquo;re responsible for maintaining a robust and fault-free production deployment. It covers monitoring fine grained parameters of a CloudStack installation and how to measure and check not just discrete values but also routinely performed actions against their expected outcome. It also dwells upon how the same discrete values make a difference from an access perspective. Whether you want to auto-resolve common problems, provide an SLA or just keep a tab on your deployment, this talk will definitely provide some insights. - - - - Does Open Source Participation Get Easier Over Time? - https://elek.github.io/acah2021/sessions/1115.html - Tue, 10 Sep 2019 17:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1115.html - What’s it like to be a committed member of an active open source community? What are the awesome parts? What are the challenges? Is open source for you? In this talk, presenters with a wide range of experience at Apache will share their stories and advice on open source participation. From starting out to ongoing involvement, from participating in open source as part of your job, to finding time to participate outside of your job, from finding models and mentors in an online, asynchronous communication environment to being such a mentor, from the worst to the best moments and experiences in open source, join us to hear about more than 30 years of combined experience with Apache projects. - - - - Hello World! Introducing Apache IoTDB - a Database for the Internet of Things - https://elek.github.io/acah2021/sessions/1078.html - Tue, 10 Sep 2019 17:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1078.html - Time series data management system is becoming increasingly important with the rise of the Internet of Things, especially in the industrial area. Apache IoTDB (incubating), one of the youngest projects in the Apache Incubator, is born for that!nApache IoTDB is an integrated data management engine designed for time series data, based on a file format which is able to store time series data very efficiently. It provides users specific services for data collection, storage and analysis. - - - - Kubernetes on BareMetal with Apache CloudStack - https://elek.github.io/acah2021/sessions/1007.html - Tue, 10 Sep 2019 17:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1007.html - nCloudStack currently supports Kubernetes deployment using Hypervisors, which enables PaaS on top of IaaS. However, PaaS applications having data-heavy loads are the most likely to be negatively impacted, especially when the service is oversubscribed. This is where BareMetal cloud offers a significant advantage over Hypervisors, especially for latency-sensitive workloads. This talk and demo will provide turnkey solution for integrating Kubernetes on CloudStack using BareMetal servers. - - - - Locking Down Apache Tomcat: Practical Security for Real-world Applications - https://elek.github.io/acah2021/sessions/1003.html - Tue, 10 Sep 2019 17:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1003.html - Out of the box, Apache Tomcat is quite secure. Then you need to configure it to suit your environment, connect your data sources, and deploy your applications. Those processes can potentially reduce the security of the entire system. A thorough review of your host, network, application and configuration is necessary to identify those areas where your security needs improvement. We’ll discuss each of these areas in some detail and how some simple tweaks and tools can make you and your users safer. - - - - Ozone: Evolving HDFS Scalability to new heights & built-in GDPR Compliance - https://elek.github.io/acah2021/sessions/1176.html - Tue, 10 Sep 2019 17:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1176.html - Apache Hadoop Ozone is a robust, distributed key-value object store for Hadoop with layered architecture and strong consistency. It separates the namespace management from block and node management layer, which allows users to independently scale on both axes. Ozone is interoperable with Hadoop ecosystem as it provides OzoneFS (Hadoop compatible file system API), data locality and plug-n-play deployment with HDFS as it can be installed in an existing Hadoop cluster and can share storage disks with HDFS. - - - - Processing CDN Logging Events at Scale - https://elek.github.io/acah2021/sessions/1199.html - Tue, 10 Sep 2019 17:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1199.html - &lsquo;Content Delivery Networks generate staggering amounts of event data at a rate that is exponentially increasing in line with the explosion of IP video. Managing this deluge with the usual suspects of Big Data platforms presents a formidable challenge, particularly in light of the often-competing goals of detailed troubleshooting investigation versus big picture analytics and machine learning. This presentation will explore capturing and processing events from the largest production deployment of Apache Traffic Server and Apache Traffic Control. - - - - Transaction Management on Cassandra - https://elek.github.io/acah2021/sessions/1139.html - Tue, 10 Sep 2019 17:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1139.html - Scalar DB is an open source library released under Apache 2 which realizes ACID-compliant transactions on Cassandra, without requiring any modifications to Cassandra itself. It achieves strongly-consistent, linearly scalable, and highly available transactions. This talk will present the theory and practice behind Scalar DB, as well as providing some benchmark results and use cases. - - - - Unleashing Apache Kafka and TensorFlow in Hybrid Cloud Architectures - https://elek.github.io/acah2021/sessions/1160.html - Tue, 10 Sep 2019 17:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1160.html - How can you leverage the flexibility and extreme scale in the public cloud combined with your Apache Kafka ecosystem to build scalable, mission-critical machine learning infrastructures, which span multiple public clouds or bridge your on-premise data centre to cloud? nThis talk will discuss and demo how you can leverage machine learning technologies such as TensorFlow with your Kafka deployments in public cloud to build a scalable, mission-critical machine learning infrastructure for data ingestion and processing, and model training, deployment and monitoring. - - - - Apache Pinot (incubating): Building Realtime Analytics Applications at LinkedIn Scale - https://elek.github.io/acah2021/sessions/1169.html - Tue, 10 Sep 2019 15:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1169.html - LinkedIn is the most advantageous social networking tool available to job seekers and business professionals today, with 610+ million members creating millions of posts, videos, and articles that generate tens of millions of shares, comments and likes per day. LinkedIn has leveraged this activity data to build rich interactive user-facing analytics applications like “Who Viewed My Profile”, Talent Insights, Ad Analytics and Publisher Analytics, among others. These applications are all powered by Pinot, as are internal dashboards, anomaly detection and root cause analysis platform like ThirdEye. - - - - Cassandra: Partition Management - https://elek.github.io/acah2021/sessions/1093.html - Tue, 10 Sep 2019 15:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1093.html - (Cassandra NGCC) Apache Cassandra on-disk storage model is based on immutable SSTables that are, depending on the compaction strategy, optimized for some general use cases. The main issue with optimally reading on-disk data is that a single partition data most of the time needs to be fetched from multiple partitions. None of the existing compaction strategies are optimized to provide a single file read guarantees on cold data. Partition based compaction strategy addresses that issue with effectively compacting all data that belongs to a single partition into the same SSTable. - - - - CloudStack identity federation: An overview - https://elek.github.io/acah2021/sessions/1144.html - Tue, 10 Sep 2019 15:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1144.html - Federated authentication in online services is increasingly being used, whether through public identity providers such as Facebook, Google and others, or in organizations' private identity providers. Identity federation is a technology that has come to simplify the life of the user, but that does not always simplify the life of application developers and operators. The purpose of this talk is to share a bit the challenges, benefits and problems of using federated authentication with multiple identity providers in Apache Cloudstack. - - - - Diversity, a beginners introduction - https://elek.github.io/acah2021/sessions/1071.html - Tue, 10 Sep 2019 15:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1071.html - Diversity and inclusion are becoming hot topics within our open source communities and industry, but for those of us &lsquo;playing life on easy&rsquo;, the subject and impact can be harder to &lsquo;get&rsquo;. Like many people who used to play life on easy, I&rsquo;ve had some blind spots around diversity and inclusion. I only discovered these through working on projects, communities and support at the ASF, and by moving to a new country! - - - - Improving the Observability of Cassandra, Kafka and Kubernetes applications with Prometheus and OpenTracing - https://elek.github.io/acah2021/sessions/1031.html - Tue, 10 Sep 2019 15:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1031.html - As distributed applications grow more complex, dynamic, and massively scalable, “observability” becomes more critical. Observability is the practice of using metrics, monitoring and distributed tracing to understand how a system works. In this presentation we’ll explore two complementary Open Source technologies: Prometheus for monitoring application metrics; and OpenTracing and Jaeger for distributed tracing. We’ll discover how they improve the observability of a massively scalable Anomaly Detection system - an application which is built around Apache Cassandra and Apache Kafka for the data layers, and dynamically deployed and scaled on Kubernetes, a container orchestration technology. - - - - Machine Learning applications for CloudStack - https://elek.github.io/acah2021/sessions/1108.html - Tue, 10 Sep 2019 15:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1108.html - While Machine learning and data mining has had profound impact on how we model applications and use data for better product consumption, there is scope for extending prediction algorithms to lower levels as well. Some useful applications of machine learning in ACS could be exploring better resource allocation that is aware of usage statistics, predicting faults, load balancing, etc. In this talk we will * take a broad overview of what Machine Learning/Data mining is and how it is being used in today&rsquo;s tech ecosystemn* explore ways in which we can make ACS more efficientn* discuss some recent advancements in how ML can benefit datacenters from research community - - - - Proxies, TLS, and Injecting Security Into Cloud Communication - https://elek.github.io/acah2021/sessions/1087.html - Tue, 10 Sep 2019 15:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1087.html - With the expansion of moving legacy applications into public clouds, there have been an increased need for features in proxies to inject security into legacy application deployments. In this talk, we present extensions made to the Apache Traffic Server proxy over the past two years to enable the secure deployment of legacy applications in a wider variety of network configurations. Specifically, we discuss how Traffic Sever can be used for Layer 4 Routing. - - - - Self-Service IoT Data Analytics with StreamPipes - https://elek.github.io/acah2021/sessions/1077.html - Tue, 10 Sep 2019 15:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1077.html - In many application domains such as manufacturing, the integration and continuous processing of real-time sensor data from the Internet of Things (IoT) is crucial to continuously monitor and detect upcoming situations. While continuous processing of events in scalable architectures is already well supported by the existing Big Data tool landscape, building such applications requires technical effort and programming skills, which is often not present in manufacturing companies. To solve this problem, we have developed StreamPipes (https://www. - - - - Serving Clean URLs via Custom Internet Media Types on Tomcat - https://elek.github.io/acah2021/sessions/1145.html - Tue, 10 Sep 2019 15:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1145.html - Increasingly static site generators are being used to create a presence on the web, both for the speed and simplicity of the resulting static pages. A related trend is the use of &lsquo;clean&rsquo; URLs with no resource name extensions, such as /products/mousetrap instead of /products/mousetrap.html. Clean URLs are more memorable, less fragile, do not leak implementation information, and are more search engine friendly. Because the default Tomcat implementation determines MIME type by filename extension, &lsquo;clean&rsquo; filenames are traditionally produced by servlet mapping or URL rewriting, perhaps coupled with Apache HTTP Server in front of Tomcat using mod_rewrite. - - - - Wire Encryption In HDFS: Protect Your Data From Others, Not Yourself - https://elek.github.io/acah2021/sessions/1175.html - Tue, 10 Sep 2019 15:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1175.html - Wire data encryption is a key component of the Hadoop Distributed File System (HDFS). HDFS can enforce different levels of data protection, allowing users to specify one based on their own needs. However, such enforcement comes in as an all-or-nothing feature. Namely, wire encryption is enforced either for all accesses or none. Since encryption bears a considerable performance cost, the all-or-nothing condition forces users to choose between &lsquo;faster but unencrypted&rsquo; or &lsquo;encrypted but slower&rsquo; for all clients. - - - - Apache Hudi (Incubating) : the past, present and future of efficient data lake architectures - https://elek.github.io/acah2021/sessions/1168.html - Tue, 10 Sep 2019 14:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1168.html - Apache Hudi is a newly incubating project at the ASF. Originally created at Uber, to power it’s vast big data lake, Hudi provides key features like atomic writes, snapshot isolation, incremental views, rollbacks, point-in-time restores, file size management &amp; many more. To date, the big data community has been polarized amongst batch and streaming systems, when balancing data freshness and scale. Hudi also addresses a combined need for speed and scale, that does not naturally fit into existing batch and streaming data processing architectures, by way of ability to support continuous ingestion and asynchronous compaction of row &amp; columnar data. - - - - Behind the Scenes of the Apache Software Foundation - https://elek.github.io/acah2021/sessions/1090.html - Tue, 10 Sep 2019 14:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1090.html - This presentation will give you everything you always wanted to know about the Apache Software Foundation (ASF), but were afraid to ask. It will show you that there is more than just the Apache web server, and provide you with information on how the ASF works and its history. The difference between membership and committership, who decides what, how elections take place, the technical infrastructure, project management committees, and the philosophy behind the incubator. - - - - Customizing virtual machines with cloud init - https://elek.github.io/acah2021/sessions/1009.html - Tue, 10 Sep 2019 14:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1009.html - This presentation will cover the basic design around how CloudStack implements support for cloud-init datasources, configuring the cloud-init service within a VM template to talk to one of the CloudStack supported datasources, and give a basic introduction to using cloud-init to customize virtual machines upon deployment. - - - - Deploying a Java-based Tomcat application to Azure - https://elek.github.io/acah2021/sessions/1159.html - Tue, 10 Sep 2019 14:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1159.html - In this talk, you will learn how you can use Azure Pipelines to deploy a Java web application to Apache Tomcat with a MySQL database on Azure. - - - - First Step to Hybrid Cloud Computation: Elastic YARN and Kubernetes - https://elek.github.io/acah2021/sessions/1174.html - Tue, 10 Sep 2019 14:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1174.html - &lsquo;In this presentation, we will introduce our work: Elastic YARN, which is an elastic, cloud-native solution and it allows co-location YARN and Kubernetes for both on-premises and on-cloud. Audiences can learn how to build a hybrid, low-cost and easy-to-use computation platform while ensuring the SLA by leveraging our work. Abstract Enterprise continues to evolve towards data-driven, more types of workloads will come onboard, the challenges of managing mixed workloads emerge. It&rsquo;s essential to think about what is the most flexible architecture to manage batch jobs and long-running services both on-premises and on-cloud. - - - - How eBay build Data Quality Platform with Apache Griffin - https://elek.github.io/acah2021/sessions/1030.html - Tue, 10 Sep 2019 14:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1030.html - High Quality data plays an important role for incubating awesome big data applications. In real projects, governing data, observing their quality health, and taking remedy actions are tedious and time-consuming. Apache Griffin is an open source Data Quality solution for Big Data, which supports both batch and streaming mode. It offers a unified process to measure your data quality from different perspectives, helping you build trusted data assets, therefore boost your confidence for your business. - - - - IoT Applications and Patterns using Apache Spark & Apache Bahir - https://elek.github.io/acah2021/sessions/1085.html - Tue, 10 Sep 2019 14:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1085.html - The Internet of Things (IoT) is all about connected devices that produce and exchange data, and building applications that produce insights from these high volumes of data are very challenging and require an understanding of multiple protocols, platforms, and other components. On this session, we will start by providing a quick introduction to IoT, some of the common analytic patterns used on IoT, and also touch on the MQTT protocol and how it is used by IoT solutions some of the quality of services tradeoffs to be considered when building an IoT application. - - - - Monitoring of CloudStack - https://elek.github.io/acah2021/sessions/1053.html - Tue, 10 Sep 2019 14:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1053.html - Our approach of monitoring the CloudStack management server, parts of CloudStacks ecosystem and its underlying components. We give an introduction of how to monitor virtual routers, customer systems and customer resources in CloudStack. - - - - Panel discussion on recognizing non-code contributions - https://elek.github.io/acah2021/sessions/1032.html - Tue, 10 Sep 2019 14:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1032.html - A discussion panel with 2 or 3 projects who have been successful at growing their communities outside the English speaking countries and how they did it. - - - - RocksDB Based Storage Engine Recent Improvements - https://elek.github.io/acah2021/sessions/1128.html - Tue, 10 Sep 2019 14:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1128.html - In last year&rsquo;s NGCC session &lsquo;Cassandra Pluggable Storage Engine&rsquo;, Instagram Cassandra team presented a new RocksDB based Cassandra storage engine which brings us huge performance win. Since then our team have been actively working on the new storage engine for adding missing features and further improving efficiency. In this session I will talk about our work on following sections and their impacts to our production system.n* Large disaggregated flash disk supportn* RocksDB based streaming improvementsn* Decoding improvementsn* Space amplification improvementsn* Some missing features added - - - - Taming the Herd - Tuning ATS for linear video - https://elek.github.io/acah2021/sessions/1114.html - Tue, 10 Sep 2019 14:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1114.html - At Comcast we have a unique challenge of serving both static and live linear content via our CDN. In general a thundering herd (multiple instantaneous simultaneous requests for the same content) is not a large concern for static files barring a few situations like serving automatic pushed update files. However when serving live content the nature of it means you are almost always generating a herd or on the verge of it due to timing. - - - - Backup and Recovery in CloudStack (it's Here!) - https://elek.github.io/acah2021/sessions/1137.html - Tue, 10 Sep 2019 12:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1137.html - This presentation will demonstrate the long awaited Backup &amp; Recovery Framework feature. The presentation will feature a real-world demonstration as well explain the design philosophy and how the feature has been built to be agnostic of the backend Backup and Recovery software in use. - - - - Ceph with CloudStack - https://elek.github.io/acah2021/sessions/1134.html - Tue, 10 Sep 2019 12:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1134.html - In this presentation, we are going to give a brief introduction to Ceph and cover some considerations around it&rsquo;s architecture, both in general and related to CloudStack. We are going to cover Ceph&rsquo;s integration into CloudStack, compare it&rsquo;s feature set versus other Primary Storage solutions for CloudStack and also share some general advises on it&rsquo;s setup. - - - - Hatching the Clutch - A Guide to the Apache Incubator - https://elek.github.io/acah2021/sessions/1113.html - Tue, 10 Sep 2019 12:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1113.html - What is the Apache Incubator? What is required from projects in order to graduate? What resources are available and which are required? What does the Foundation provide? Why is the dev email list so important? What about your Mentors? We are all volunteers and strive to be your gateway into taking full advantage of what the Foundation has to offer. Let&rsquo;s review the state of the Clutch and see what is happening now. - - - - Let's Encrypt Apache Tomcat - https://elek.github.io/acah2021/sessions/1001.html - Tue, 10 Sep 2019 12:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1001.html - The Let&rsquo;s Encrypt initiative has changed the landscape for online security. The capital cost of cryptographic security for every web site worldwide has been lowered from potentially thousands of dollars per year to zero. One of the trade-offs for the availability of these free certificates is that they are short-lived, making automated deployment of certificates an essential piece of any deployment plan. Let&rsquo;s Encrypt supports Apache httpd out of the box for single-command renewal and deployment on a single host, while with Tomcat it&rsquo;s more complicated. - - - - Managing Hundreds of Petabytes of data in the Cloud - https://elek.github.io/acah2021/sessions/1012.html - Tue, 10 Sep 2019 12:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1012.html - Twitter&rsquo;s on premise HDFS clusters are scaled out to handle hundreds of petabyte data serving thousands of developers every day. Managing such large scale data introduces several challenges. Organization, discovery, data life cycle, replication, schema management and many more questions needs to be answered at scale. Our data infrastructure is built on different open source and in house projects working together to solve this problem. While we extend our data to cloud, we had to solve similar challenges on cloud environment at Twitter scale. - - - - Next Generation Cassandra Compaction, Going beyond LCS - https://elek.github.io/acah2021/sessions/1011.html - Tue, 10 Sep 2019 12:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1011.html - Cassandra users with general purpose workloads typically must choose the lesser of two evils between the default “write optimized” Size Tiered Compaction Strategy (STCS) and the “read optimized” Leveled Compaction Strategy (LCS). It is typical these days to hear “use LCS unless it explodes, in which case use STCS…unless that explodes, in which case you are out of luck”. I believe that by taking what we as a community learned from the development and real world deployment of LCS, we can create a next generation compaction strategy which offers all of the benefits of LCS with few of the drawbacks. - - - - Observability in distributed system - https://elek.github.io/acah2021/sessions/1129.html - Tue, 10 Sep 2019 12:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1129.html - Observability is quite important for the cloud native application. Itncould be a challenge if we build up the system observer from thenstrache. By leveraging the Skywalking&rsquo;s extension mechanism, we couldnsend out the data to the backend directly to improve thenObservability of the Cloud Native Service. In distributed applications, horizontal sharding and transactio coordination are most popular solutions. Observability for businessninvocation on sharding scenario are not enough; but observer allnsharded details of backend infrastructures indifferently are tooncomplicated. - - - - Optimizing Big Data Pipelines with Apache Nemo (Incubating) - https://elek.github.io/acah2021/sessions/1167.html - Tue, 10 Sep 2019 12:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1167.html - Optimizing scheduling and communication behaviors of big data pipelines for resource and data characteristics is crucial for achieving high performance. For example, optimizing for geographically-distributed resources, cheap transient resources, disk-based large data shuffle, and skewed data have recently received a lot of interest and attention. The new incubating Apache Nemo project aims to make it easy to express and enforce such optimizations by providing a policy interface that transforms an intermediate representation (IR) of data processing applications. - - - - Peeking Behind the Curtain: The Anatomy of a Real Major Incident - https://elek.github.io/acah2021/sessions/1212.html - Tue, 10 Sep 2019 12:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1212.html - Failures are inevitable. But when they occur, our goal should be to resolve them as quickly and efficiently as possible. PagerDuty has developed an open-source Incident Response framework based on the Incident Command System (ICS). That product-independent process has helped many organizations set up Incident Response processes that resolve technical issues as quickly and effectively as possible. There&rsquo;s a lot of documentation you can follow, but how do incidents actually play out in real-time when they happen? - - - - Processing IoT Data from End to End with MQTT and Apache Kafka - https://elek.github.io/acah2021/sessions/1107.html - Tue, 10 Sep 2019 12:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1107.html - This session discusses end-to-end use cases such as connected cars, smart home or healthcare sensors where you integrate IoT devices with enterprise IT using open source technologies and standards. MQTT is a lightweight messaging protocol for Internet of Things (IoT). However, MQTT is not built for high scalability, longer storage or easy integration to legacy systems. Apache Kafka is a highly scalable distributed streaming platform, which ingests, stores, processes and forwards high volumes of data from thousands of IoT devices. - - - - Traffic Server Remap Update - https://elek.github.io/acah2021/sessions/1064.html - Tue, 10 Sep 2019 12:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1064.html - Rewriting URLs, or &lsquo;remapping&rsquo;, is probably the single most used feature of the Traffic Servernproxy. With the transition to YAML from a proprietary line based format, it was decided to startnover with effectively a blank slate, to design a configuration with a functionality emphasizingnpower and flexibility over compatibility. This talk will elaborate on the decision path leading upnto this complete restructuring, the design of the configuration, and examples of its use. - - - - Apache Cassandra community health - https://elek.github.io/acah2021/sessions/1076.html - Tue, 10 Sep 2019 11:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1076.html - This talk will discuss the current health of the Apache Cassandra community looking at everything from community adoption, user adoption, health of contributors/committers, project velocity and a number of other measures of community health. In this talk Ben will also discuss broader market changes and challenges that impact the Apache Cassandra community drawn from our experience at Instaclustr, discussions with industry analysts and public data sources. This talk will be an update on the state of Apache Cassandra community presentation from NGCC 2017. - - - - Apache Iceberg: a table format for distributed databases - https://elek.github.io/acah2021/sessions/1166.html - Tue, 10 Sep 2019 11:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1166.html - &lsquo;This talk will introduce Apache Iceberg (Incubating), a new table format designed for distributed databases that can be used directly by any size application. It will cover Iceberg’s design and the features that distinguish it from other table formats, including: -Reliable schema evolutio -Hidden partitioningn-Atomic operationsn-Support for object stores like S3n-Support for direct client access This talk will also present how Iceberg performs at scale in Netflix’s environment.&rsquo; - - - - Apache SkyWalking -- Top level open source APM - https://elek.github.io/acah2021/sessions/1058.html - Tue, 10 Sep 2019 11:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1058.html - SkyWalking has just graduated as Top Level Project in this year. SkyWalking is an open source APM system, including monitoring, tracing, diagnosing capabilities for distributed systems in Cloud Native architecture. In the modern world, we have legacy and cloud-native distributed systems, even in hybrid. SkyWalking provides different observability methods, including language agents and service mesh angle, which adopt different scenarios. In this session, I will introduce what SkyWalking do, and how we do it, even without requiring any big data tech. - - - - How Netflix manages petabyte scale Apache Cassandra in the cloud - https://elek.github.io/acah2021/sessions/1010.html - Tue, 10 Sep 2019 11:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1010.html - At Netflix, we manage petabytes of data in Apache Cassandra which must be reliably accessible to users in mere milliseconds. To achieve this, we have built sophisticated control planes that turn our persistence layer based on Apache Cassandra into a truly self-driving system. We will start with the user interface that Netflix developers use to interact with their Cassandra databases and dive deep into the automation that powers it all. From cluster creation, through scaling up, to cluster death, complex automation drives large fleets of virtual machines hosted on the AWS cloud. - - - - Lessons learned while migrating 3000 openvz containers to CloudStack - https://elek.github.io/acah2021/sessions/1127.html - Tue, 10 Sep 2019 11:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1127.html - When Leaseweb acquired a new company 3000 public cloud customers we had to come up with a strategy to move all these VMs to our CloudStack setup. During this talk we will share: how we migrated a large foreign openvz environment into our kvm cloudstack environment in an automated fashion minimizing customer impact. The lessons we learned along the way and how we would do it again in the future. - - - - Life beyond Kafka with Apache Pulsar - https://elek.github.io/acah2021/sessions/1153.html - Tue, 10 Sep 2019 11:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1153.html - In a world where everything is connected, trillions of devices are sending data in real-time to data centers. Apache Kafka as a streaming platform is the system by excellence for many years. However, there are many competitors who are making its life more difficult. In this talk, I will introduce to one of them, Apache Pulsar sponsored by Yahoo itself. Pulsar is the next generation messaging system that uses a different architecture to achieve durability, performance, scalability, efficiency, multi-tenancy and geographic replication. - - - - State of the cat, new and upcoming - https://elek.github.io/acah2021/sessions/1035.html - Tue, 10 Sep 2019 11:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1035.html - This presentation will detail the current state of the Apache Tomcat project and its community. After going through information on recent releases, community metrics and security advisories, this will cover important new and upcoming features. This includes HTTP/X support, clustering, TLS, cloud enablement, and much more. - - - - The Apache Way - Not only the What but the Why - https://elek.github.io/acah2021/sessions/1056.html - Tue, 10 Sep 2019 11:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1056.html - The Apache Way consists of the basic core tenets, embraced by all ASF projects, which define how Open Source projects should be managed, governed and maintained. But in addition to a simply list of what those tenets are, we will dig deeper and discover the history and rationale behind them all. The incredible success of the Apache Way can be used to immediate advantage in your open source project and actual forms the core philosophy behind InnerSource. - - - - The future of the CloudStack Virtual Router - https://elek.github.io/acah2021/sessions/1157.html - Tue, 10 Sep 2019 11:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1157.html - This talk surveys and explores the scopes and design for the next-generation network model consolidation in CloudStack and a new super-fast micro-vm based VR. The talk will discuss the scope and goals for the next-generation VR, consolidation of basic and advanced networks and zones, flexible networking topologies, UI based network designing, a standard systemvm patching mechanism and uniform VR programming interface with a lightweight secured agent, faster rules programming and zero-downtime upgrades. - - - - Using Apache Traffic Control to cache any web object at scale - https://elek.github.io/acah2021/sessions/1002.html - Tue, 10 Sep 2019 11:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1002.html - This talk will focus on how any organization can deploy a Content Delivery Network (CDN) to cache web object at scale using Apache Traffic Control (ATC) and Apache Traffic Server (ATS). Topics covered will be a brief introduction to the projects, a high-level overview of traffic routing on a CDN, followed by a deep dive into how to efficiently cache objects on an ATC CDN using ATS. Based on years of experience operating ATC CDNs, several different scenarios with corresponding object types (video, web content, etc) will be presented. - - - - YuniKorn: A Universal Resource Scheduler for both YARN and Kubernetes - https://elek.github.io/acah2021/sessions/1173.html - Tue, 10 Sep 2019 11:15:00 +0000 - - https://elek.github.io/acah2021/sessions/1173.html - &lsquo;We will talk about our open source work - YuniKorn scheduler project (Y for YARN, K for K8s, uni- for Unified) brings long-wanted features such as hierarchical queues, fairness between users/jobs/queues, preemption to Kubernetes; and it brings service scheduling enhancements to YARN. Any improvements to this scheduler can benefit both Kubernetes and YARN community. Abstract: YARN schedulers are optimized for high-throughput, multi-tenant batch workloads. It can scale up to 50k nodes per cluster, and schedule 20k containers per second; On the other side, Kubernetes schedulers are optimized for long-running services, but many features like hierarchical queues, fairness resource sharing, and preemption etc, are either missing or not mature enough at this point of time. - - - - Evaluating Content/Text Extraction at Scale with Apache Tika - https://elek.github.io/acah2021/sessions/1072.html - Mon, 09 Sep 2019 17:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1072.html - Text extraction tools are essential for obtaining the textual content and metadata of computer files for use in a wide variety of applications, including search and natural language processing tools. Techniques and tools for evaluating text extraction tools are largely missing from academia and industry. This talk will focus on recent improvements to Apache Tika&rsquo;s tika-eval module to help integrators evaluate content extraction at scale. The tika-eval module was initially developed for a single batch mode on a single vm. - - - - Network Visibility in Hybrid and Multi Clouds - https://elek.github.io/acah2021/sessions/1037.html - Mon, 09 Sep 2019 17:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1037.html - Network Visibility in Apache CloudStack - - - - Realtime Geospatial Analytics with GPUs, RAPIDS, and Apache Arrow - https://elek.github.io/acah2021/sessions/1165.html - Mon, 09 Sep 2019 17:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1165.html - Prior to RAPIDS, geospatial analytics, especially networking analytics for routing, required large CPU clusters to process. Even with 100s of machines, it would take hours if not days to get answers. With RAPIDS, and Apache 2.0 open source project built on Apache Arrow, graph analytics, clustering, and many other geospatial workflows can be completed end to end in seconds. Learn how to load data (CSV, Parquet, or ORC) directly into GPU memory with cuIO, process it with Dask-cuDF, and analyze it with cuML and cuGraph in seconds, all on a single node. - - - - There's life in the (old) dog yet: Apache Edgent reinvents itself and Edge Analytics - https://elek.github.io/acah2021/sessions/1016.html - Mon, 09 Sep 2019 17:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1016.html - Apache Edgent (formerly Quarks) is a Project in the Apache Incubator since early 2016. The main focus of the project is to simplify edge processing pipelines. Through the last year it went a bit silent around the project and although the topic is hotter than ever the project lost its drive and went into hibernation. With the rise of other projects in the (industrial) IoT ecosystem like Apache PLC4X and Apache IoTDB there are a lot of new use cases and applications for edge processing, especially in the field of industrial IoT thus, a group of contributors from other IoT related projects has gathered to bring the project back to life and on track. - - - - Workshop: Getting Apache Traffic Control Up And Running - https://elek.github.io/acah2021/sessions/1046.html - Mon, 09 Sep 2019 17:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1046.html - CDN-in-a-Box is a subproject within Apache Traffic Control (ATC) to build a test/experimental CDN in docker on an individual workstation. The process is still challenging, so I&rsquo;d like to do a workshop to guide interested users through the process on their own laptops. The goal is to show them how the CDN works at a fairly low level and how to test that the various parts work as expected as well as troubleshooting if things go awry. - - - - A Mile in My Shoes: How Empathy & Compassion Can Transform Your Work & Your Life - https://elek.github.io/acah2021/sessions/1200.html - Mon, 09 Sep 2019 16:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1200.html - In many places, we&rsquo;re told to &lsquo;leave our emotions out of it,&rsquo; in our work, and in our other technical endeavors, including in open source communities. But that is an irrational, dangerously naïve point of view, one that often leads us to not concern ourselves with the feelings of others, thereby allowing biases (intended or not) to come into play. This harms our communities, by excluding valuable contributors, and their skills and ideas. - - - - Cloudstack API development 101 - https://elek.github.io/acah2021/sessions/1039.html - Mon, 09 Sep 2019 16:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1039.html - This is a primer to getting started with adding functionality to CloudStack. In this presentation we will go over a custom Maven archetype that will fast track CloudStack development, whether the goal is to develop functionality to contribute to CloudStack project or to integrate CloudStack with site specific systems. It is geared toward Java developers who are new to the CloudStack code base. - - - - Customer Cloud Portal - https://elek.github.io/acah2021/sessions/1162.html - Mon, 09 Sep 2019 16:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1162.html - A technical overview showing how we at itelligence as a cloud service provider use CloudStack and how we extent our infrastructure with additional service components to enhance our cloud service portfolio. - - - - GeoSpatial and Temporal Forecasting in Uber Marketplace - https://elek.github.io/acah2021/sessions/1061.html - Mon, 09 Sep 2019 16:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1061.html - Uber’s Marketplace is the algorithmic brain and decision engine behind our ride-sharing services. Marketplace Forecasting builds and deploys ML algorithms to handle the immense coordination, hyperlocal decision making, and learning needed to tackle the enormous scale and movement of our transportation network. In order for our decision engines to be future-aware, we need to be able to “see into the future” as accurately as possible across both space and time. In Uber, we use H3 (a hexagonal hierarchical geospatial indexing system) to partition the data geospatially. - - - - Machine Learning and Data Analytics in the Content Delivery Context - https://elek.github.io/acah2021/sessions/1063.html - Mon, 09 Sep 2019 16:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1063.html - In line with the content delivery community&rsquo;s interest in data-driven operational insights and improving operational efficiency, this presentation and discussion will address the following topics: (1) What are Machine Learning, Deep Learning, and the different types of analytics (from descriptive through prescriptive)? (2) What are the potential benefits and risks of analytical solutions? (3) And finally - a walkthrough of some example use cases from modern content delivery networks and their corresponding analytical solutions. - - - - The First Mile -- Edge and IoT Data Collection with Apache NiFi and MiNiFi - https://elek.github.io/acah2021/sessions/1118.html - Mon, 09 Sep 2019 16:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1118.html - Apache NiFi provided a revolutionary data flow management system with a broad range of integrations with existing data production, consumption, and analysis ecosystems, all covered with robust data delivery and provenance infrastructure. Now learn about the follow-on project which expands the reach of NiFi to the edge, Apache MiNiFi. MiNiFi is a lightweight application which can be deployed on hardware orders of magnitude smaller and less powerful than the existing standard data collection platforms. - - - - Anycast ATC Integration at Comcast - https://elek.github.io/acah2021/sessions/1047.html - Mon, 09 Sep 2019 15:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1047.html - Anycast is an industry proven technology and is quickly becoming a requirement for off-the-shelf CDNs as well as CDN-as-a-Service. In this talk we will investigate how Comcast is integrating anycast functionality into Apache Traffic Control to improve resiliency, convergence and localization. - - - - Cloudstack Locking Service - https://elek.github.io/acah2021/sessions/3001.html - Mon, 09 Sep 2019 15:00:00 +0000 - - https://elek.github.io/acah2021/sessions/3001.html - As CloudStack cannot work with any mysql clustering solution, it is time to explore a new locking service, manager and pluggable interface which would allow CloudStack DB to be HA enabled with multi-master read/write. Talk will focus on, - Need for a locking service and challenges with existing CloudStack architecture - Different possible clustering solution that can be adopted - Showcasing a PoC for future implementation with minimal changes to existing architecture using percona xtradb or any other clustering solution - Additionally, explore the idea of getting rid of mshost table, and use locking service to find about other management servers. - - - - Geospatial Indexing and Search at Scale with Apache Lucene - https://elek.github.io/acah2021/sessions/1000.html - Mon, 09 Sep 2019 15:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1000.html - Come have a look under the covers at new data structures that enable geospatial and multi-dimensional indexing and search at massive scale in Apache Lucene. This talk will cover the indexing structures considered and ultimately implemented in the Apache Lucene Open Source Project along with the 25 - 30X boost in performance and centimeter spatial accuracy achieved in the latest release. Join us and see what&rsquo;s next for scalable Geospatial Search in Apache Lucene. - - - - Going Beyond StringUtils: A Broader Look at Apache Commons - https://elek.github.io/acah2021/sessions/1152.html - Mon, 09 Sep 2019 15:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1152.html - The Apache Commons Project is home to one of the more widely consumed Java utility functions, namely “StringUtils.isBlank(String str);”. But when you look beyond commons-lang, you’ll find a wide array of heavily utilized and extremely useful tools. For example, commons-cli is the command line parser used by the Apache Maven project; commons-fileupload is heavily used by both Apache Tomcat and Apache Struts; commons-text is heavily used by Apache Hadoop; and commons-dbcp is a database connection pool used across Tomcat as well as the Spring Framework. - - - - State of Storage in Cloudstack - https://elek.github.io/acah2021/sessions/1100.html - Mon, 09 Sep 2019 15:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1100.html - With the increased focus and legislation around data sovereignty requirements, storage solutions play an increasingly important role in cloud infrastructures. We will present an overview of the evolution of storage orchestration in Apache CloudStack, with a focus on managed storage. It will also shed light on some upcoming features and the future direction. - - - - Using Apache Pulsar to Provide Real-Time IoT Analytics on the Edge - https://elek.github.io/acah2021/sessions/1015.html - Mon, 09 Sep 2019 15:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1015.html - he business value of data decreases rapidly after it is created, particularly in use cases such as fraud prevention, cybersecurity, and real-time system monitoring. The high-volume, high-velocity datasets used to feed these use cases often contain valuable, but perishable, insights that must be acted upon immediately. In order to maximize the value of their data enterprises must fundamentally change their approach to processing real-time data to focusing reducing their decision latency on the perishable insights that exist within their real-time data streams. - - - - Apache Commons State of the Union - https://elek.github.io/acah2021/sessions/1154.html - Mon, 09 Sep 2019 14:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1154.html - We will go over the state of the Apache Commons Project. Where things stand, and will take a look at the future road map for the project. - - - - GeoMesa on top of Accumulo, HBase, Cassandra, and big data file formats for massive geospatial data - a LocationTech Project - https://elek.github.io/acah2021/sessions/1164.html - Mon, 09 Sep 2019 14:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1164.html - LocationTech is the geospatial software working group of the Eclipse Foundation. The projects range from fundamental libraries that provide spatial operations to complex library suites which coordinate multiple Apache projects to build complete spatial processing solutions.nLocationTech GeoMesa builds on top of distributed Apache databases like Accumulo, HBase, Cassandra and Kafka to provide indexing, querying, and analysis for large spatio-temporal datasets. GeoMesa does this by integrating other LocationTech projects like JTS, Spatial4J, and SFCurve with these databases as well as Apache open source file formats such as Avro, Arrow, Orc, and Parquet. - - - - Let's Encrypt Apache Traffic Control - https://elek.github.io/acah2021/sessions/1198.html - Mon, 09 Sep 2019 14:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1198.html - Let&rsquo;s Encrypt is a free, automated certificate authority (CA) which is open for use to anyone who owns a domain name. Using Traffic Router as the DNS, we have integrated with Let&rsquo;s Encrypt when utilizing HTTPS on the delivery service domains. Many open source projects provide libraries to ease implementation of Let&rsquo;s Encrypt, and for this project we used go-acme/lego (MIT licensed) to integrate with the Traffic Ops Golang API. Integration with Let&rsquo;s Encrypt provides many advantages including saving money by using a completely free CA, saving time with automation including certificate renewal, and improving reliability by eliminating manual intervention for certificate generation and deployment. - - - - Open vSwitch with DPDK on CloudStack - https://elek.github.io/acah2021/sessions/1040.html - Mon, 09 Sep 2019 14:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1040.html - Support for Open vSwitch with DPDK has been introduced on CloudStack version 4.12. In this talk we will go through the motivations and advantages of using DPDK on KVM hosts, high level architecture of the feature, what is needed and how can it be used on production environments and future works. - - - - Troubleshooting the Virtual Router - Run and Get Diagnostics - https://elek.github.io/acah2021/sessions/1098.html - Mon, 09 Sep 2019 14:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1098.html - Demonstration of CloudStack’s latest features for troubleshooting the Virtual Router connectivity and configuration, called Run and Get Diagnostics. Run Diagnostics gives the admin ability to execute diagnostics commands native the VR OS directly from the CloudStack UI. Admin is able to determine connectivity capabilities of the VR without logging into the SystemVM at all. Get diagnostics feature allow the admin to gather information about the SystemVM as fast as clicking on a button in the ‘QuickView’ of the VM. - - - - Using PLC4X and LF Edge EVE to build a next generation Industry 4.0 IoT Platform - https://elek.github.io/acah2021/sessions/1131.html - Mon, 09 Sep 2019 14:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1131.html - Industrial IoT platforms are undergoing rapid shift away from vertically integrated systems built mostly out of embedded software towards an Edge Virtualization Platforms and commodity, cloud native software. While the cloud, big-data, fast-data, machine-learning has been a huge thing in the IT industry for the last years, the production industry was stuck in a proprietary world, dominated by some of the big players. Two open source projects now offer the building blocks for the new platform that can hope to break that proprietary chokehold. - - - - Apache Science Data Analytics Platform (SDAP) - https://elek.github.io/acah2021/sessions/1163.html - Mon, 09 Sep 2019 11:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1163.html - An Analytics Center Framework (ACF) is an environment that enables the confluence of resources for scientific investigation. It harmonizes data, tools and computational resources which subsequently enable the research community to focus on the investigation. The Earth science community is an innovative community. We produce many tools and solutions to improve how we do science. In computer science, a framework is a reusable, semi-complete application that can be specialized to produce custom applications [Johnson:88]. - - - - Automating Apache Traffic Control Environment Deployments with Ansible - https://elek.github.io/acah2021/sessions/1062.html - Mon, 09 Sep 2019 11:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1062.html - Apache Traffic Control is a set of applications designed to complement Apache Traffic Server to comprise a Content Delivery Network. Currently the creation of a complete production-ready CDN environment is a complex process. I will be demonstrating automation leveraging Ansible and CDN environment designs to facilitate consistent and more rapid testing environment creation. - - - - Building a smart query mediation engine for Genome Research using Apache Drill - https://elek.github.io/acah2021/sessions/1051.html - Mon, 09 Sep 2019 11:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1051.html - The integration of both clinical and genomics data into a single platformnthat will allow the researchers to search, analyze and discover the hidde patterns in biomedical databases. Developing this integrated platform be-ncomes a complex and challenging task due to (i) heterogeneous data typesn(ii) huge volume of data (iii) number of attributes are di erent across thendata (iv) generated data are from heterogeneous platform and (v) corre-nlation between clinical and genomics data are di erent to each other andn(vi) scalability factors. - - - - Episode IV - A New Hope: Industry 4.0 done our way - https://elek.github.io/acah2021/sessions/1079.html - Mon, 09 Sep 2019 11:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1079.html - It is a period of technological change. Rebel Industry 4.0 open-source software solutions, striking from hidden bases, have won their first victories against the evil proprietary Industrial Empire. During the battle, Rebel spies managed to implement industrial protocols to the Empire’s ultimate weapons, the industrial Programmable Logic Controllers, almost unprotected pieces of hardware with enough power to destroy an entire planet. Pursued by the Empire’s sinister agents, a representative of the project races to Las Vegas aboard his intercontinental sub-orbital shuttle, custodian of his plans to revolutionize the way we are building Industry 4. - - - - NoVNC console implementation in CloudStack - https://elek.github.io/acah2021/sessions/1109.html - Mon, 09 Sep 2019 11:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1109.html - NoVNC is an open-source browser based VNC client, implemented using HTML5 technologies (Web Sockets, Canvas) with encryption (wss://) support. Proposed NoVNC implementation provides better performance and is more responsive when compared to the current ajax-based console in Apache CloudStack. Our implementation is based on NoVNC and websockify. Websockify is a websocket proxy converting WebSocket request from browser to TCP request and then forward it to VNC server. We have already implemented a NoVNC console for CloudStack 4. - - - - Openstack from a CloudStack perspective - https://elek.github.io/acah2021/sessions/1097.html - Mon, 09 Sep 2019 11:30:00 +0000 - - https://elek.github.io/acah2021/sessions/1097.html - An experienced CloudStack developer will show his journey to deploy an OpenStack cloud. We will compare the pros and cons of CloudStack and OpenStack from the following perspectives:n(i) Deployment: cost, time and risks of deploying a CloudStack vs OpenStack infrastructure;n(ii) Maintenance: the costs and time required on maintaining a Cloud orchestrated with CloudStack and another with OpenStack; and,n(ii) Development: how development on both orchestrators is done? What looks like the codebase of each project and which technologies are involved? - - - - Apache Mynewt - ASF's Embedded OS Project - https://elek.github.io/acah2021/sessions/1105.html - Mon, 09 Sep 2019 10:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1105.html - Apache Mynewt is an operating system for low powered embedded systems based on 32-bit microcontrollers. At the core, it is a small pre-emptive RTOS and an extensive set of modules for connectivity, file systems, power and performance management options and lots more. It is a highly flexible, responsive, general purpose OS for constrained devices, think of it as embedded Linux for devices that can’t run Linux. Features include a secure bootloader, image management, hardware abstraction, instrumentation for stats and logs. - - - - Automating Apache Traffic Control Server Deployments with IPv6 - https://elek.github.io/acah2021/sessions/1020.html - Mon, 09 Sep 2019 10:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1020.html - Apache Traffic Control has some built-in tools for assisting in the installation and deployment of new servers; however, it is still a somewhat manual and repetitive process. I will be demonstrating how the Traffic Control API, used in conjunction with IPv6 autoconf can be used to streamline the deployment process, allowing servers to configure themselves post-install. Will highlight the use of the ‘tc-netconfig’ application, which was developed to facilitate this process. - - - - Customized CloudStack in Leaseweb - https://elek.github.io/acah2021/sessions/1161.html - Mon, 09 Sep 2019 10:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1161.html - In this talk we will share how Leaseweb customized the CloudStack project to serve our business model and our customers, by doing some improvements and do some bug fixes. We always focused on upgrading our fork to the latest in the community while making sure all our improvements are ported to the latest version. In this talk we will tell you about some features and fixes we did like: - multiple NICs feature, - NoVNC, autoscaling, - SSL offloading in load balancer - and other changes made to serve some of our customers' needs. - - - - Geospatial Data Management in Apache Spark - https://elek.github.io/acah2021/sessions/1117.html - Mon, 09 Sep 2019 10:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1117.html - The volume of spatial data increases at a staggering rate. This talk comprehensively studies how existing works, such as GeoSpark, extend Apache Spark to uphold massive-scale spatial data. During this talk, we first provide a background introduction of the characteristics of spatial data and the history of distributed data management systems. A follow-up section presents the common approaches used by the practitioners to extend Spark and introduces the vital components in a generic spatial data management system. - - - - Java + Serverless - Finally a match made in Heaven? - https://elek.github.io/acah2021/sessions/1214.html - Mon, 09 Sep 2019 10:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1214.html - Why isn’t Java a more popular language for serverless developers? How does Apache Openwhisk work with Java? Despite there being 10 million Java developers, Java is being used for only 6% of serverless workloads according to a recent survey. Historical problems with long cold-starts, heavy-weight frameworks and lack of tooling have made Java an unpopular choice for serverless development…. until now! In this session, I will take you on a journey to explain what has changed with Java to finally make it an amazing language for building serverless applications, using the open source Apache Openwhisk platform! - - - - Modern UI for CloudStack - https://elek.github.io/acah2021/sessions/1143.html - Mon, 09 Sep 2019 10:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1143.html - The CloudStack UI code has tolerated massive technical debt over the years and the UI codebase has reached a breaking point. In this talk, we will do a comparison of modern UI development techniques, the pros and cons of current UI UX and codebase, ideas and experiments of a new UI for CloudStack. We need to discuss and propose a new UI that:n* Has CloudMonkey like api discovery and show/hide/block, buttons/sections based on what is allowedn* URL router to navigate to resources, and clicking back button does not close the UIn* Easy to maintain, implement and extend via plugins etc. - - - - From (Single) ISP-CDN - To Multi-ISP CDN - https://elek.github.io/acah2021/sessions/1197.html - Mon, 09 Sep 2019 09:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1197.html - &lsquo;Traffic-control was built as an ISP CDN, serving content within a single ISP. Using traffic-control for creating a multi-ISP CDN, puts up challenges in various aspects: starting from the deployment of the different TC components outside of the ISPs, towards the isolation between different ISPs, as well as self-service, dealing simultaneously with different structures of ISP networks, the changing popularity of content in different ISP etc. In this session we would like to described few of the above challenges, as well as the derived traffic-control components requirements, flexibility and modularity. - - - - Geospatial Data and Processing - Reusable Building Blocks - https://elek.github.io/acah2021/sessions/1150.html - Mon, 09 Sep 2019 09:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1150.html - Reuse of common elements for geospatial information and processing results in increased productivity, lower interoperability friction, and higher data quality. This presentation provides a survey of reusable geospatial building blocks. Common practices for coordinate reference systems (CRSs), spatial geometries and data arrays used for projects with geospatial content will be described based on open source projects and open standards. Emphasis is placed on the use of open standards including the recently updated OGC CRS Well Known Text (CRS WKT) and OGC APIs. - - - - How to Become an IoT Developer (and Have Fun!) - https://elek.github.io/acah2021/sessions/1120.html - Mon, 09 Sep 2019 09:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1120.html - I started off my life as a developer writing machine code and C and working on some low-level hardware projects. Then this thing called the internet come along and I moved into the web application space for a couple of decades. More recently I&rsquo;ve moved back into commercial IoT development and not unexpectedly a lot has changed over that time. In this talk, I&rsquo;ll cover what it&rsquo;s like developing IoT projects. - - - - New Features of Apache RocketMQ - https://elek.github.io/acah2021/sessions/1213.html - Mon, 09 Sep 2019 09:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1213.html - In the past year, Apache RocketMQ has released a series of new features including transactional messages, ACLs, message track, multiple replicas storage, and multi-language clients. These features are all from contributors from many companies. nThis talk is not only about bringing technical analysis related to these new features, but also will share how to collaborate on the development of important features in the community, which is very important for Apache. - - - - Welcome to Cloudstack Collaboration Conference - https://elek.github.io/acah2021/sessions/1112.html - Mon, 09 Sep 2019 09:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1112.html - The CloudStack Project VP welcomes delegates to the CloudStack Collaboration Conference. And gives opening remarks about the CloudStack community, with special guest joining on stage. - - - - Welcome to the Content Delivery Track - https://elek.github.io/acah2021/sessions/1045.html - Mon, 09 Sep 2019 08:00:00 +0000 - - https://elek.github.io/acah2021/sessions/1045.html - This presentation will open the track and serve as a welcome to those attending. I will cover what’s new in the ATC (Apache Traffic Control) community, introduce our new committers and PMC members, talk about what we will cover during the conference and then talk a little about the CDN market as a whole. - - - - Apache Cassandra - https://elek.github.io/acah2021/tracks/cassandra.html - Mon, 01 Jan 0001 00:00:00 +0000 - - https://elek.github.io/acah2021/tracks/cassandra.html - - - - - Apache Drill - https://elek.github.io/acah2021/tracks/drill.html - Mon, 01 Jan 0001 00:00:00 +0000 - - https://elek.github.io/acah2021/tracks/drill.html - - - - - Apache Karaf - https://elek.github.io/acah2021/tracks/karaf.html - Mon, 01 Jan 0001 00:00:00 +0000 - - https://elek.github.io/acah2021/tracks/karaf.html - - - - - Apache Tomcat - https://elek.github.io/acah2021/tracks/tomcat.html - Mon, 01 Jan 0001 00:00:00 +0000 - - https://elek.github.io/acah2021/tracks/tomcat.html - - - - - ApacheCon Code of Conduct - https://elek.github.io/acah2021/conduct.html - Mon, 01 Jan 0001 00:00:00 +0000 - - https://elek.github.io/acah2021/conduct.html - At the Apache Software Foundation, all online interactions are governed by the ASF Code of Conduct, while in-person events are subject to the Anti-Harassment Policy. -ApacheCon is dedicated to providing a harassment-free experience for everyone. We do not tolerate harassment of participants in any form. Participants violating these rules may be sanctioned or expelled without a refund, at the discretion of the event organizers. -Harassment includes offensive verbal comments, deliberate intimidation, stalking, following, unwanted photography or recording, sustained disruption of talks or other events, inappropriate physical contact, and unwelcome sexual attention. - - - - Beam Summit - https://elek.github.io/acah2021/tracks/beam.html - Mon, 01 Jan 0001 00:00:00 +0000 - - https://elek.github.io/acah2021/tracks/beam.html - - - - - Big Data - https://elek.github.io/acah2021/tracks/bigdata.html - Mon, 01 Jan 0001 00:00:00 +0000 - - https://elek.github.io/acah2021/tracks/bigdata.html - - - - - Call for Presentations - https://elek.github.io/acah2021/cfp.html - Mon, 01 Jan 0001 00:00:00 +0000 - - https://elek.github.io/acah2021/cfp.html - The call for presentations is now open! Submit proposals by Monday, May 3rd, 2021 8:00 AM (America/New_York time - UTC -5). Please do not wait until the last minute. -Submit your talk proposal now! -We are looking for presentations about anything relating to Apache Software Foundation projects, open source governance, community, and software development. -In particular, this year we are building content tracks around the following specific topics/projects: - API/Microservice Cassandra Content Delivery Community Drill Geospatial Incubating IoT Observability Royale Tomcat Big Data Community Content Management Fineract/Fintech Groovy Integration Karaf RDF/Linked Data Search Social Data If your proposed presentation falls into one of these categories, please select that topic in the CfP entry form. - - - - Cloudstack Collaboration Conference - https://elek.github.io/acah2021/tracks/cloudstack.html - Mon, 01 Jan 0001 00:00:00 +0000 - - https://elek.github.io/acah2021/tracks/cloudstack.html - - - - - Community - https://elek.github.io/acah2021/tracks/community.html - Mon, 01 Jan 0001 00:00:00 +0000 - - https://elek.github.io/acah2021/tracks/community.html - - - - - Content Delivery - https://elek.github.io/acah2021/tracks/contentdelivery.html - Mon, 01 Jan 0001 00:00:00 +0000 - - https://elek.github.io/acah2021/tracks/contentdelivery.html - - - - - General - https://elek.github.io/acah2021/tracks/general.html - Mon, 01 Jan 0001 00:00:00 +0000 - - https://elek.github.io/acah2021/tracks/general.html - - - - - Geospatial software - https://elek.github.io/acah2021/tracks/geospatial.html - Mon, 01 Jan 0001 00:00:00 +0000 - - https://elek.github.io/acah2021/tracks/geospatial.html - - - - - Graph processing - https://elek.github.io/acah2021/tracks/graphprocessing.html - Mon, 01 Jan 0001 00:00:00 +0000 - - https://elek.github.io/acah2021/tracks/graphprocessing.html - - - - - Integration - https://elek.github.io/acah2021/tracks/integration.html - Mon, 01 Jan 0001 00:00:00 +0000 - - https://elek.github.io/acah2021/tracks/integration.html - - - - - Internet of Things - https://elek.github.io/acah2021/tracks/internetofthings.html - Mon, 01 Jan 0001 00:00:00 +0000 - - https://elek.github.io/acah2021/tracks/internetofthings.html - - - - - Machine Learning - https://elek.github.io/acah2021/tracks/machinelearning.html - Mon, 01 Jan 0001 00:00:00 +0000 - - https://elek.github.io/acah2021/tracks/machinelearning.html - - - - - Mobile - https://elek.github.io/acah2021/tracks/mobile.html - Mon, 01 Jan 0001 00:00:00 +0000 - - https://elek.github.io/acah2021/tracks/mobile.html - - - - - Observability - https://elek.github.io/acah2021/tracks/observability.html - Mon, 01 Jan 0001 00:00:00 +0000 - - https://elek.github.io/acah2021/tracks/observability.html - - - - - OFBiz - https://elek.github.io/acah2021/tracks/ofbiz.html - Mon, 01 Jan 0001 00:00:00 +0000 - - https://elek.github.io/acah2021/tracks/ofbiz.html - - - - - Streaming - https://elek.github.io/acah2021/tracks/streaming.html - Mon, 01 Jan 0001 00:00:00 +0000 - - https://elek.github.io/acah2021/tracks/streaming.html - - - - - diff --git a/docs/js/init.js b/docs/js/init.js deleted file mode 100644 index b28377a08..000000000 --- a/docs/js/init.js +++ /dev/null @@ -1,83 +0,0 @@ -/* - Solarize by TEMPLATED - templated.co @templatedco - Released for free under the Creative Commons Attribution 3.0 license (templated.co/license) -*/ - -(function($) { - - skel.init({ - reset: 'full', - breakpoints: { - global: { range: '*', href: 'css/style.css', containers: 1400, grid: { gutters: 50 } }, - wide: { range: '-1680', href: 'css/style-wide.css', containers: 1200, grid: { gutters: 40 } }, - normal: { range: '-1280', href: 'css/style-normal.css', containers: 960, lockViewport: true }, - narrow: { range: '-980', href: 'css/style-narrow.css', containers: '95%', grid: { gutters: 25 } }, - narrower: { range: '-840', href: 'css/style-narrower.css', grid: { collapse: true } }, - mobile: { range: '-640', href: 'css/style-mobile.css', containers: '90%', grid: { gutters: 15 } } - } - }, { - layers: { - layers: { - navPanel: { - animation: 'pushX', - breakpoints: 'narrower', - clickToClose: true, - height: '100%', - hidden: true, - html: '
', - orientation: 'vertical', - position: 'top-left', - side: 'left', - width: 275 - }, - titleBar: { - breakpoints: 'narrower', - height: 44, - html: '', - position: 'top-left', - side: 'top', - width: '100%' - } - } - } - }); - - $(function() { - - var $window = $(window), - $body = $('body'); - - // Disable animations/transitions until page has loaded. - $body.addClass('loading'); - - $window.on('load', function() { - $body.removeClass('loading'); - }); - - // Forms (IE<10). - var $form = $('form'); - if ($form.length > 0) { - - $form.find('.form-button-submit') - .on('click', function() { - $(this).parents('form').submit(); - return false; - }); - - if (skel.vars.IEVersion < 10) { - $.fn.n33_formerize=function(){var _fakes=new Array(),_form = $(this);_form.find('input[type=text],textarea').each(function() { var e = $(this); if (e.val() == '' || e.val() == e.attr('placeholder')) { e.addClass('formerize-placeholder'); e.val(e.attr('placeholder')); } }).blur(function() { var e = $(this); if (e.attr('name').match(/_fakeformerizefield$/)) return; if (e.val() == '') { e.addClass('formerize-placeholder'); e.val(e.attr('placeholder')); } }).focus(function() { var e = $(this); if (e.attr('name').match(/_fakeformerizefield$/)) return; if (e.val() == e.attr('placeholder')) { e.removeClass('formerize-placeholder'); e.val(''); } }); _form.find('input[type=password]').each(function() { var e = $(this); var x = $($('
').append(e.clone()).remove().html().replace(/type="password"/i, 'type="text"').replace(/type=password/i, 'type=text')); if (e.attr('id') != '') x.attr('id', e.attr('id') + '_fakeformerizefield'); if (e.attr('name') != '') x.attr('name', e.attr('name') + '_fakeformerizefield'); x.addClass('formerize-placeholder').val(x.attr('placeholder')).insertAfter(e); if (e.val() == '') e.hide(); else x.hide(); e.blur(function(event) { event.preventDefault(); var e = $(this); var x = e.parent().find('input[name=' + e.attr('name') + '_fakeformerizefield]'); if (e.val() == '') { e.hide(); x.show(); } }); x.focus(function(event) { event.preventDefault(); var x = $(this); var e = x.parent().find('input[name=' + x.attr('name').replace('_fakeformerizefield', '') + ']'); x.hide(); e.show().focus(); }); x.keypress(function(event) { event.preventDefault(); x.val(''); }); }); _form.submit(function() { $(this).find('input[type=text],input[type=password],textarea').each(function(event) { var e = $(this); if (e.attr('name').match(/_fakeformerizefield$/)) e.attr('name', ''); if (e.val() == e.attr('placeholder')) { e.removeClass('formerize-placeholder'); e.val(''); } }); }).bind("reset", function(event) { event.preventDefault(); $(this).find('select').val($('option:first').val()); $(this).find('input,textarea').each(function() { var e = $(this); var x; e.removeClass('formerize-placeholder'); switch (this.type) { case 'submit': case 'reset': break; case 'password': e.val(e.attr('defaultValue')); x = e.parent().find('input[name=' + e.attr('name') + '_fakeformerizefield]'); if (e.val() == '') { e.hide(); x.show(); } else { e.show(); x.hide(); } break; case 'checkbox': case 'radio': e.attr('checked', e.attr('defaultValue')); break; case 'text': case 'textarea': e.val(e.attr('defaultValue')); if (e.val() == '') { e.addClass('formerize-placeholder'); e.val(e.attr('placeholder')); } break; default: e.val(e.attr('defaultValue')); break; } }); window.setTimeout(function() { for (x in _fakes) _fakes[x].trigger('formerize_sync'); }, 10); }); return _form; }; - $form.n33_formerize(); - } - - } - - // Dropdowns. - $('#nav > ul').dropotron({ - offsetY: -15, - hoverDelay: 0 - }); - - }); - -})(jQuery); diff --git a/docs/js/jquery.dropotron.min.js b/docs/js/jquery.dropotron.min.js deleted file mode 100644 index 20ebb39ae..000000000 --- a/docs/js/jquery.dropotron.min.js +++ /dev/null @@ -1,2 +0,0 @@ -/* jquery.dropotron.js v1.4.2 | (c) n33 | n33.co | MIT licensed */ -(function(e){var t="openerActiveClass",n="left",r="doCollapseAll",i="position",s="trigger",o="disableSelection_dropotron",u="addClass",a="doCollapse",f=!1,l="outerWidth",c="removeClass",h="preventDefault",p="dropotron",d="clearTimeout",v="length",m="right",g="speed",y=!0,b="parent",w="none",E="stopPropagation",S=":visible",x="absolute",T="click",N="doExpand",C="css",k="center",L="toggle",A="baseZIndex",O="offsetX",M="alignment",_="children",D="submenuClassPrefix",P="doToggle",H="hover",B="ul",j="relative",F="opacity",I="z-index",q="opener",R="find",U="px",z=null,W="fadeTo",X="offset";e.fn[o]=function(){return e(this)[C]("user-select",w)[C]("-khtml-user-select",w)[C]("-moz-user-select",w)[C]("-o-user-select",w)[C]("-webkit-user-select",w)},e.fn[p]=function(t){var n;if(this[v]>1)for(n=0;n0&&r.add(p).on("mouseleave",function(){window[d](ut),ut=window.setTimeout(function(){r[s](a)},et.hideDelay)}),r[o]().hide()[u](et.menuClass)[C](i,x).on("mouseenter",function(){window[d](ut)}).on(N,function(){var o,h,v,w,E,T,N,_,D,P,H;if(r.is(S))return f;window[d](ut),nt.each(function(){var t=e(this);e.contains(t.get(0),p.get(0))||t[s](a)}),o=p[X](),h=p[i](),v=p[b]()[i](),w=p[l](),E=r[l](),T=r[C](I)==et[A];if(T){et.detach?N=o:N=h,P=N.top+p.outerHeight()+et.globalOffsetY,_=et[M],r[c](n)[c](m)[c](k);switch(et[M]){case m:D=N[n]-E+w,D<0&&(D=N[n],_=n);break;case k:D=N[n]-Math.floor((E-w)/2),D<0?(D=N[n],_=n):D+E>it.width()&&(D=N[n]-E+w,_=m);break;case n:default:D=N[n],D+E>it.width()&&(D=N[n]-E+w,_=m)}r[u](_)}else{p[C](i)==j||p[C](i)==x?(P=et.offsetY,D=-1*h[n]):(P=h.top+et.offsetY,D=0);switch(et[M]){case m:D+=-1*p[b]()[l]()+et[O];break;case k:case n:default:D+=p[b]()[l]()+et[O]}}navigator.userAgent.match(/MSIE ([0-9]+)\./)&&RegExp.$1<8&&(D+=et.IEOffsetX,P+=et.IEOffsetY),r[C](n,D+U)[C]("top",P+U),r[C](F,"0.01").show(),H=f,p[C](i)==j||p[C](i)==x?D=-1*h[n]:D=0,r[X]()[n]<0?(D+=p[b]()[l]()-et[O],H=y):r[X]()[n]+E>it.width()&&(D+=-1*p[b]()[l]()-et[O],H=y),H&&r[C](n,D+U),r.hide()[C](F,"1");switch(et.mode){case"zoom":st=y,p[u](et[t]),r.animate({width:L,height:L},et[g],et.easing,function(){st=f});break;case"slide":st=y,p[u](et[t]),r.animate({height:L},et[g],et.easing,function(){st=f});break;case"fade":st=y,T&&!et.noOpenerFade?(et[g]=="slow"?H=80:et[g]=="fast"?H=40:H=Math.floor(et[g]/2),p[W](H,.01,function(){p[u](et[t]),p[W](et[g],1),r.fadeIn(et[g],function(){st=f})})):(p[u](et[t]),p[W](et[g],1),r.fadeIn(et[g],function(){st=f}));break;case"instant":default:p[u](et[t]),r.show()}return f}).on(a,function(){return r.is(S)?(r.hide(),p[c](et[t]),r[R]("."+et[t])[c](et[t]),r[R](B).hide(),f):f}).on(P,function(){return r.is(S)?r[s](a):r[s](N),f}),p[o]()[u](q)[C]("cursor","pointer").on(T,function(e){if(st)return;e[h](),e[E](),r[s](P)}),et.expandMode==H&&p[H](function(){if(st)return;ot=window.setTimeout(function(){r[s](N)},et.hoverDelay)},function(){window[d](ot)})}),nt[R]("a")[C]("display","block").on(T,function(t){if(st)return;e(this).attr("href")[v]<1&&t[h]()}),tt[R]("li")[C]("white-space","nowrap").each(function(){var t=e(this),n=t[_]("a"),i=t[_](B);n.on(T,function(t){e(this).attr("href")[v]<1?t[h]():t[E]()}),n[v]>0&&i[v]==0&&t.on(T,function(e){if(st)return;tt[s](r),e[E]()})}),tt[_]("li").each(function(){var t,n,r,i,s=e(this),o=s[_](B);if(o[v]>0){et.detach&&(et.cloneOnDetach&&(t=o.clone(),t.attr("class","").hide().appendTo(o[b]())),o.detach().appendTo(rt));for(n=et[A],r=1,i=o;i[v]>0;r++)i[C](I,n++),et[D]&&i[u](et[D]+(n-1-et[A])),i=i[R]("> li > ul")}}),it.on("scroll",function(){tt[s](r)}).on("keypress",function(e){!st&&e.keyCode==27&&(e[h](),tt[s](r))}),rt.on(T,function(){st||tt[s](r)})}})(jQuery); \ No newline at end of file diff --git a/docs/js/jquery.min.js b/docs/js/jquery.min.js deleted file mode 100644 index 73f33fb3a..000000000 --- a/docs/js/jquery.min.js +++ /dev/null @@ -1,4 +0,0 @@ -/*! jQuery v1.11.0 | (c) 2005, 2014 jQuery Foundation, Inc. | jquery.org/license */ -!function(a,b){"object"==typeof module&&"object"==typeof module.exports?module.exports=a.document?b(a,!0):function(a){if(!a.document)throw new Error("jQuery requires a window with a document");return b(a)}:b(a)}("undefined"!=typeof window?window:this,function(a,b){var c=[],d=c.slice,e=c.concat,f=c.push,g=c.indexOf,h={},i=h.toString,j=h.hasOwnProperty,k="".trim,l={},m="1.11.0",n=function(a,b){return new n.fn.init(a,b)},o=/^[\s\uFEFF\xA0]+|[\s\uFEFF\xA0]+$/g,p=/^-ms-/,q=/-([\da-z])/gi,r=function(a,b){return b.toUpperCase()};n.fn=n.prototype={jquery:m,constructor:n,selector:"",length:0,toArray:function(){return d.call(this)},get:function(a){return null!=a?0>a?this[a+this.length]:this[a]:d.call(this)},pushStack:function(a){var b=n.merge(this.constructor(),a);return b.prevObject=this,b.context=this.context,b},each:function(a,b){return n.each(this,a,b)},map:function(a){return this.pushStack(n.map(this,function(b,c){return a.call(b,c,b)}))},slice:function(){return this.pushStack(d.apply(this,arguments))},first:function(){return this.eq(0)},last:function(){return this.eq(-1)},eq:function(a){var b=this.length,c=+a+(0>a?b:0);return this.pushStack(c>=0&&b>c?[this[c]]:[])},end:function(){return this.prevObject||this.constructor(null)},push:f,sort:c.sort,splice:c.splice},n.extend=n.fn.extend=function(){var a,b,c,d,e,f,g=arguments[0]||{},h=1,i=arguments.length,j=!1;for("boolean"==typeof g&&(j=g,g=arguments[h]||{},h++),"object"==typeof g||n.isFunction(g)||(g={}),h===i&&(g=this,h--);i>h;h++)if(null!=(e=arguments[h]))for(d in e)a=g[d],c=e[d],g!==c&&(j&&c&&(n.isPlainObject(c)||(b=n.isArray(c)))?(b?(b=!1,f=a&&n.isArray(a)?a:[]):f=a&&n.isPlainObject(a)?a:{},g[d]=n.extend(j,f,c)):void 0!==c&&(g[d]=c));return g},n.extend({expando:"jQuery"+(m+Math.random()).replace(/\D/g,""),isReady:!0,error:function(a){throw new Error(a)},noop:function(){},isFunction:function(a){return"function"===n.type(a)},isArray:Array.isArray||function(a){return"array"===n.type(a)},isWindow:function(a){return null!=a&&a==a.window},isNumeric:function(a){return a-parseFloat(a)>=0},isEmptyObject:function(a){var b;for(b in a)return!1;return!0},isPlainObject:function(a){var b;if(!a||"object"!==n.type(a)||a.nodeType||n.isWindow(a))return!1;try{if(a.constructor&&!j.call(a,"constructor")&&!j.call(a.constructor.prototype,"isPrototypeOf"))return!1}catch(c){return!1}if(l.ownLast)for(b in a)return j.call(a,b);for(b in a);return void 0===b||j.call(a,b)},type:function(a){return null==a?a+"":"object"==typeof a||"function"==typeof a?h[i.call(a)]||"object":typeof a},globalEval:function(b){b&&n.trim(b)&&(a.execScript||function(b){a.eval.call(a,b)})(b)},camelCase:function(a){return a.replace(p,"ms-").replace(q,r)},nodeName:function(a,b){return a.nodeName&&a.nodeName.toLowerCase()===b.toLowerCase()},each:function(a,b,c){var d,e=0,f=a.length,g=s(a);if(c){if(g){for(;f>e;e++)if(d=b.apply(a[e],c),d===!1)break}else for(e in a)if(d=b.apply(a[e],c),d===!1)break}else if(g){for(;f>e;e++)if(d=b.call(a[e],e,a[e]),d===!1)break}else for(e in a)if(d=b.call(a[e],e,a[e]),d===!1)break;return a},trim:k&&!k.call("\ufeff\xa0")?function(a){return null==a?"":k.call(a)}:function(a){return null==a?"":(a+"").replace(o,"")},makeArray:function(a,b){var c=b||[];return null!=a&&(s(Object(a))?n.merge(c,"string"==typeof a?[a]:a):f.call(c,a)),c},inArray:function(a,b,c){var d;if(b){if(g)return g.call(b,a,c);for(d=b.length,c=c?0>c?Math.max(0,d+c):c:0;d>c;c++)if(c in b&&b[c]===a)return c}return-1},merge:function(a,b){var c=+b.length,d=0,e=a.length;while(c>d)a[e++]=b[d++];if(c!==c)while(void 0!==b[d])a[e++]=b[d++];return a.length=e,a},grep:function(a,b,c){for(var d,e=[],f=0,g=a.length,h=!c;g>f;f++)d=!b(a[f],f),d!==h&&e.push(a[f]);return e},map:function(a,b,c){var d,f=0,g=a.length,h=s(a),i=[];if(h)for(;g>f;f++)d=b(a[f],f,c),null!=d&&i.push(d);else for(f in a)d=b(a[f],f,c),null!=d&&i.push(d);return e.apply([],i)},guid:1,proxy:function(a,b){var c,e,f;return"string"==typeof b&&(f=a[b],b=a,a=f),n.isFunction(a)?(c=d.call(arguments,2),e=function(){return a.apply(b||this,c.concat(d.call(arguments)))},e.guid=a.guid=a.guid||n.guid++,e):void 0},now:function(){return+new Date},support:l}),n.each("Boolean Number String Function Array Date RegExp Object Error".split(" "),function(a,b){h["[object "+b+"]"]=b.toLowerCase()});function s(a){var b=a.length,c=n.type(a);return"function"===c||n.isWindow(a)?!1:1===a.nodeType&&b?!0:"array"===c||0===b||"number"==typeof b&&b>0&&b-1 in a}var t=function(a){var b,c,d,e,f,g,h,i,j,k,l,m,n,o,p,q,r,s="sizzle"+-new Date,t=a.document,u=0,v=0,w=eb(),x=eb(),y=eb(),z=function(a,b){return a===b&&(j=!0),0},A="undefined",B=1<<31,C={}.hasOwnProperty,D=[],E=D.pop,F=D.push,G=D.push,H=D.slice,I=D.indexOf||function(a){for(var b=0,c=this.length;c>b;b++)if(this[b]===a)return b;return-1},J="checked|selected|async|autofocus|autoplay|controls|defer|disabled|hidden|ismap|loop|multiple|open|readonly|required|scoped",K="[\\x20\\t\\r\\n\\f]",L="(?:\\\\.|[\\w-]|[^\\x00-\\xa0])+",M=L.replace("w","w#"),N="\\["+K+"*("+L+")"+K+"*(?:([*^$|!~]?=)"+K+"*(?:(['\"])((?:\\\\.|[^\\\\])*?)\\3|("+M+")|)|)"+K+"*\\]",O=":("+L+")(?:\\(((['\"])((?:\\\\.|[^\\\\])*?)\\3|((?:\\\\.|[^\\\\()[\\]]|"+N.replace(3,8)+")*)|.*)\\)|)",P=new RegExp("^"+K+"+|((?:^|[^\\\\])(?:\\\\.)*)"+K+"+$","g"),Q=new RegExp("^"+K+"*,"+K+"*"),R=new RegExp("^"+K+"*([>+~]|"+K+")"+K+"*"),S=new RegExp("="+K+"*([^\\]'\"]*?)"+K+"*\\]","g"),T=new RegExp(O),U=new RegExp("^"+M+"$"),V={ID:new RegExp("^#("+L+")"),CLASS:new RegExp("^\\.("+L+")"),TAG:new RegExp("^("+L.replace("w","w*")+")"),ATTR:new RegExp("^"+N),PSEUDO:new RegExp("^"+O),CHILD:new RegExp("^:(only|first|last|nth|nth-last)-(child|of-type)(?:\\("+K+"*(even|odd|(([+-]|)(\\d*)n|)"+K+"*(?:([+-]|)"+K+"*(\\d+)|))"+K+"*\\)|)","i"),bool:new RegExp("^(?:"+J+")$","i"),needsContext:new RegExp("^"+K+"*[>+~]|:(even|odd|eq|gt|lt|nth|first|last)(?:\\("+K+"*((?:-\\d)?\\d*)"+K+"*\\)|)(?=[^-]|$)","i")},W=/^(?:input|select|textarea|button)$/i,X=/^h\d$/i,Y=/^[^{]+\{\s*\[native \w/,Z=/^(?:#([\w-]+)|(\w+)|\.([\w-]+))$/,$=/[+~]/,_=/'|\\/g,ab=new RegExp("\\\\([\\da-f]{1,6}"+K+"?|("+K+")|.)","ig"),bb=function(a,b,c){var d="0x"+b-65536;return d!==d||c?b:0>d?String.fromCharCode(d+65536):String.fromCharCode(d>>10|55296,1023&d|56320)};try{G.apply(D=H.call(t.childNodes),t.childNodes),D[t.childNodes.length].nodeType}catch(cb){G={apply:D.length?function(a,b){F.apply(a,H.call(b))}:function(a,b){var c=a.length,d=0;while(a[c++]=b[d++]);a.length=c-1}}}function db(a,b,d,e){var f,g,h,i,j,m,p,q,u,v;if((b?b.ownerDocument||b:t)!==l&&k(b),b=b||l,d=d||[],!a||"string"!=typeof a)return d;if(1!==(i=b.nodeType)&&9!==i)return[];if(n&&!e){if(f=Z.exec(a))if(h=f[1]){if(9===i){if(g=b.getElementById(h),!g||!g.parentNode)return d;if(g.id===h)return d.push(g),d}else if(b.ownerDocument&&(g=b.ownerDocument.getElementById(h))&&r(b,g)&&g.id===h)return d.push(g),d}else{if(f[2])return G.apply(d,b.getElementsByTagName(a)),d;if((h=f[3])&&c.getElementsByClassName&&b.getElementsByClassName)return G.apply(d,b.getElementsByClassName(h)),d}if(c.qsa&&(!o||!o.test(a))){if(q=p=s,u=b,v=9===i&&a,1===i&&"object"!==b.nodeName.toLowerCase()){m=ob(a),(p=b.getAttribute("id"))?q=p.replace(_,"\\$&"):b.setAttribute("id",q),q="[id='"+q+"'] ",j=m.length;while(j--)m[j]=q+pb(m[j]);u=$.test(a)&&mb(b.parentNode)||b,v=m.join(",")}if(v)try{return G.apply(d,u.querySelectorAll(v)),d}catch(w){}finally{p||b.removeAttribute("id")}}}return xb(a.replace(P,"$1"),b,d,e)}function eb(){var a=[];function b(c,e){return a.push(c+" ")>d.cacheLength&&delete b[a.shift()],b[c+" "]=e}return b}function fb(a){return a[s]=!0,a}function gb(a){var b=l.createElement("div");try{return!!a(b)}catch(c){return!1}finally{b.parentNode&&b.parentNode.removeChild(b),b=null}}function hb(a,b){var c=a.split("|"),e=a.length;while(e--)d.attrHandle[c[e]]=b}function ib(a,b){var c=b&&a,d=c&&1===a.nodeType&&1===b.nodeType&&(~b.sourceIndex||B)-(~a.sourceIndex||B);if(d)return d;if(c)while(c=c.nextSibling)if(c===b)return-1;return a?1:-1}function jb(a){return function(b){var c=b.nodeName.toLowerCase();return"input"===c&&b.type===a}}function kb(a){return function(b){var c=b.nodeName.toLowerCase();return("input"===c||"button"===c)&&b.type===a}}function lb(a){return fb(function(b){return b=+b,fb(function(c,d){var e,f=a([],c.length,b),g=f.length;while(g--)c[e=f[g]]&&(c[e]=!(d[e]=c[e]))})})}function mb(a){return a&&typeof a.getElementsByTagName!==A&&a}c=db.support={},f=db.isXML=function(a){var b=a&&(a.ownerDocument||a).documentElement;return b?"HTML"!==b.nodeName:!1},k=db.setDocument=function(a){var b,e=a?a.ownerDocument||a:t,g=e.defaultView;return e!==l&&9===e.nodeType&&e.documentElement?(l=e,m=e.documentElement,n=!f(e),g&&g!==g.top&&(g.addEventListener?g.addEventListener("unload",function(){k()},!1):g.attachEvent&&g.attachEvent("onunload",function(){k()})),c.attributes=gb(function(a){return a.className="i",!a.getAttribute("className")}),c.getElementsByTagName=gb(function(a){return a.appendChild(e.createComment("")),!a.getElementsByTagName("*").length}),c.getElementsByClassName=Y.test(e.getElementsByClassName)&&gb(function(a){return a.innerHTML="
",a.firstChild.className="i",2===a.getElementsByClassName("i").length}),c.getById=gb(function(a){return m.appendChild(a).id=s,!e.getElementsByName||!e.getElementsByName(s).length}),c.getById?(d.find.ID=function(a,b){if(typeof b.getElementById!==A&&n){var c=b.getElementById(a);return c&&c.parentNode?[c]:[]}},d.filter.ID=function(a){var b=a.replace(ab,bb);return function(a){return a.getAttribute("id")===b}}):(delete d.find.ID,d.filter.ID=function(a){var b=a.replace(ab,bb);return function(a){var c=typeof a.getAttributeNode!==A&&a.getAttributeNode("id");return c&&c.value===b}}),d.find.TAG=c.getElementsByTagName?function(a,b){return typeof b.getElementsByTagName!==A?b.getElementsByTagName(a):void 0}:function(a,b){var c,d=[],e=0,f=b.getElementsByTagName(a);if("*"===a){while(c=f[e++])1===c.nodeType&&d.push(c);return d}return f},d.find.CLASS=c.getElementsByClassName&&function(a,b){return typeof b.getElementsByClassName!==A&&n?b.getElementsByClassName(a):void 0},p=[],o=[],(c.qsa=Y.test(e.querySelectorAll))&&(gb(function(a){a.innerHTML="",a.querySelectorAll("[t^='']").length&&o.push("[*^$]="+K+"*(?:''|\"\")"),a.querySelectorAll("[selected]").length||o.push("\\["+K+"*(?:value|"+J+")"),a.querySelectorAll(":checked").length||o.push(":checked")}),gb(function(a){var b=e.createElement("input");b.setAttribute("type","hidden"),a.appendChild(b).setAttribute("name","D"),a.querySelectorAll("[name=d]").length&&o.push("name"+K+"*[*^$|!~]?="),a.querySelectorAll(":enabled").length||o.push(":enabled",":disabled"),a.querySelectorAll("*,:x"),o.push(",.*:")})),(c.matchesSelector=Y.test(q=m.webkitMatchesSelector||m.mozMatchesSelector||m.oMatchesSelector||m.msMatchesSelector))&&gb(function(a){c.disconnectedMatch=q.call(a,"div"),q.call(a,"[s!='']:x"),p.push("!=",O)}),o=o.length&&new RegExp(o.join("|")),p=p.length&&new RegExp(p.join("|")),b=Y.test(m.compareDocumentPosition),r=b||Y.test(m.contains)?function(a,b){var c=9===a.nodeType?a.documentElement:a,d=b&&b.parentNode;return a===d||!(!d||1!==d.nodeType||!(c.contains?c.contains(d):a.compareDocumentPosition&&16&a.compareDocumentPosition(d)))}:function(a,b){if(b)while(b=b.parentNode)if(b===a)return!0;return!1},z=b?function(a,b){if(a===b)return j=!0,0;var d=!a.compareDocumentPosition-!b.compareDocumentPosition;return d?d:(d=(a.ownerDocument||a)===(b.ownerDocument||b)?a.compareDocumentPosition(b):1,1&d||!c.sortDetached&&b.compareDocumentPosition(a)===d?a===e||a.ownerDocument===t&&r(t,a)?-1:b===e||b.ownerDocument===t&&r(t,b)?1:i?I.call(i,a)-I.call(i,b):0:4&d?-1:1)}:function(a,b){if(a===b)return j=!0,0;var c,d=0,f=a.parentNode,g=b.parentNode,h=[a],k=[b];if(!f||!g)return a===e?-1:b===e?1:f?-1:g?1:i?I.call(i,a)-I.call(i,b):0;if(f===g)return ib(a,b);c=a;while(c=c.parentNode)h.unshift(c);c=b;while(c=c.parentNode)k.unshift(c);while(h[d]===k[d])d++;return d?ib(h[d],k[d]):h[d]===t?-1:k[d]===t?1:0},e):l},db.matches=function(a,b){return db(a,null,null,b)},db.matchesSelector=function(a,b){if((a.ownerDocument||a)!==l&&k(a),b=b.replace(S,"='$1']"),!(!c.matchesSelector||!n||p&&p.test(b)||o&&o.test(b)))try{var d=q.call(a,b);if(d||c.disconnectedMatch||a.document&&11!==a.document.nodeType)return d}catch(e){}return db(b,l,null,[a]).length>0},db.contains=function(a,b){return(a.ownerDocument||a)!==l&&k(a),r(a,b)},db.attr=function(a,b){(a.ownerDocument||a)!==l&&k(a);var e=d.attrHandle[b.toLowerCase()],f=e&&C.call(d.attrHandle,b.toLowerCase())?e(a,b,!n):void 0;return void 0!==f?f:c.attributes||!n?a.getAttribute(b):(f=a.getAttributeNode(b))&&f.specified?f.value:null},db.error=function(a){throw new Error("Syntax error, unrecognized expression: "+a)},db.uniqueSort=function(a){var b,d=[],e=0,f=0;if(j=!c.detectDuplicates,i=!c.sortStable&&a.slice(0),a.sort(z),j){while(b=a[f++])b===a[f]&&(e=d.push(f));while(e--)a.splice(d[e],1)}return i=null,a},e=db.getText=function(a){var b,c="",d=0,f=a.nodeType;if(f){if(1===f||9===f||11===f){if("string"==typeof a.textContent)return a.textContent;for(a=a.firstChild;a;a=a.nextSibling)c+=e(a)}else if(3===f||4===f)return a.nodeValue}else while(b=a[d++])c+=e(b);return c},d=db.selectors={cacheLength:50,createPseudo:fb,match:V,attrHandle:{},find:{},relative:{">":{dir:"parentNode",first:!0}," ":{dir:"parentNode"},"+":{dir:"previousSibling",first:!0},"~":{dir:"previousSibling"}},preFilter:{ATTR:function(a){return a[1]=a[1].replace(ab,bb),a[3]=(a[4]||a[5]||"").replace(ab,bb),"~="===a[2]&&(a[3]=" "+a[3]+" "),a.slice(0,4)},CHILD:function(a){return a[1]=a[1].toLowerCase(),"nth"===a[1].slice(0,3)?(a[3]||db.error(a[0]),a[4]=+(a[4]?a[5]+(a[6]||1):2*("even"===a[3]||"odd"===a[3])),a[5]=+(a[7]+a[8]||"odd"===a[3])):a[3]&&db.error(a[0]),a},PSEUDO:function(a){var b,c=!a[5]&&a[2];return V.CHILD.test(a[0])?null:(a[3]&&void 0!==a[4]?a[2]=a[4]:c&&T.test(c)&&(b=ob(c,!0))&&(b=c.indexOf(")",c.length-b)-c.length)&&(a[0]=a[0].slice(0,b),a[2]=c.slice(0,b)),a.slice(0,3))}},filter:{TAG:function(a){var b=a.replace(ab,bb).toLowerCase();return"*"===a?function(){return!0}:function(a){return a.nodeName&&a.nodeName.toLowerCase()===b}},CLASS:function(a){var b=w[a+" "];return b||(b=new RegExp("(^|"+K+")"+a+"("+K+"|$)"))&&w(a,function(a){return b.test("string"==typeof a.className&&a.className||typeof a.getAttribute!==A&&a.getAttribute("class")||"")})},ATTR:function(a,b,c){return function(d){var e=db.attr(d,a);return null==e?"!="===b:b?(e+="","="===b?e===c:"!="===b?e!==c:"^="===b?c&&0===e.indexOf(c):"*="===b?c&&e.indexOf(c)>-1:"$="===b?c&&e.slice(-c.length)===c:"~="===b?(" "+e+" ").indexOf(c)>-1:"|="===b?e===c||e.slice(0,c.length+1)===c+"-":!1):!0}},CHILD:function(a,b,c,d,e){var f="nth"!==a.slice(0,3),g="last"!==a.slice(-4),h="of-type"===b;return 1===d&&0===e?function(a){return!!a.parentNode}:function(b,c,i){var j,k,l,m,n,o,p=f!==g?"nextSibling":"previousSibling",q=b.parentNode,r=h&&b.nodeName.toLowerCase(),t=!i&&!h;if(q){if(f){while(p){l=b;while(l=l[p])if(h?l.nodeName.toLowerCase()===r:1===l.nodeType)return!1;o=p="only"===a&&!o&&"nextSibling"}return!0}if(o=[g?q.firstChild:q.lastChild],g&&t){k=q[s]||(q[s]={}),j=k[a]||[],n=j[0]===u&&j[1],m=j[0]===u&&j[2],l=n&&q.childNodes[n];while(l=++n&&l&&l[p]||(m=n=0)||o.pop())if(1===l.nodeType&&++m&&l===b){k[a]=[u,n,m];break}}else if(t&&(j=(b[s]||(b[s]={}))[a])&&j[0]===u)m=j[1];else while(l=++n&&l&&l[p]||(m=n=0)||o.pop())if((h?l.nodeName.toLowerCase()===r:1===l.nodeType)&&++m&&(t&&((l[s]||(l[s]={}))[a]=[u,m]),l===b))break;return m-=e,m===d||m%d===0&&m/d>=0}}},PSEUDO:function(a,b){var c,e=d.pseudos[a]||d.setFilters[a.toLowerCase()]||db.error("unsupported pseudo: "+a);return e[s]?e(b):e.length>1?(c=[a,a,"",b],d.setFilters.hasOwnProperty(a.toLowerCase())?fb(function(a,c){var d,f=e(a,b),g=f.length;while(g--)d=I.call(a,f[g]),a[d]=!(c[d]=f[g])}):function(a){return e(a,0,c)}):e}},pseudos:{not:fb(function(a){var b=[],c=[],d=g(a.replace(P,"$1"));return d[s]?fb(function(a,b,c,e){var f,g=d(a,null,e,[]),h=a.length;while(h--)(f=g[h])&&(a[h]=!(b[h]=f))}):function(a,e,f){return b[0]=a,d(b,null,f,c),!c.pop()}}),has:fb(function(a){return function(b){return db(a,b).length>0}}),contains:fb(function(a){return function(b){return(b.textContent||b.innerText||e(b)).indexOf(a)>-1}}),lang:fb(function(a){return U.test(a||"")||db.error("unsupported lang: "+a),a=a.replace(ab,bb).toLowerCase(),function(b){var c;do if(c=n?b.lang:b.getAttribute("xml:lang")||b.getAttribute("lang"))return c=c.toLowerCase(),c===a||0===c.indexOf(a+"-");while((b=b.parentNode)&&1===b.nodeType);return!1}}),target:function(b){var c=a.location&&a.location.hash;return c&&c.slice(1)===b.id},root:function(a){return a===m},focus:function(a){return a===l.activeElement&&(!l.hasFocus||l.hasFocus())&&!!(a.type||a.href||~a.tabIndex)},enabled:function(a){return a.disabled===!1},disabled:function(a){return a.disabled===!0},checked:function(a){var b=a.nodeName.toLowerCase();return"input"===b&&!!a.checked||"option"===b&&!!a.selected},selected:function(a){return a.parentNode&&a.parentNode.selectedIndex,a.selected===!0},empty:function(a){for(a=a.firstChild;a;a=a.nextSibling)if(a.nodeType<6)return!1;return!0},parent:function(a){return!d.pseudos.empty(a)},header:function(a){return X.test(a.nodeName)},input:function(a){return W.test(a.nodeName)},button:function(a){var b=a.nodeName.toLowerCase();return"input"===b&&"button"===a.type||"button"===b},text:function(a){var b;return"input"===a.nodeName.toLowerCase()&&"text"===a.type&&(null==(b=a.getAttribute("type"))||"text"===b.toLowerCase())},first:lb(function(){return[0]}),last:lb(function(a,b){return[b-1]}),eq:lb(function(a,b,c){return[0>c?c+b:c]}),even:lb(function(a,b){for(var c=0;b>c;c+=2)a.push(c);return a}),odd:lb(function(a,b){for(var c=1;b>c;c+=2)a.push(c);return a}),lt:lb(function(a,b,c){for(var d=0>c?c+b:c;--d>=0;)a.push(d);return a}),gt:lb(function(a,b,c){for(var d=0>c?c+b:c;++db;b++)d+=a[b].value;return d}function qb(a,b,c){var d=b.dir,e=c&&"parentNode"===d,f=v++;return b.first?function(b,c,f){while(b=b[d])if(1===b.nodeType||e)return a(b,c,f)}:function(b,c,g){var h,i,j=[u,f];if(g){while(b=b[d])if((1===b.nodeType||e)&&a(b,c,g))return!0}else while(b=b[d])if(1===b.nodeType||e){if(i=b[s]||(b[s]={}),(h=i[d])&&h[0]===u&&h[1]===f)return j[2]=h[2];if(i[d]=j,j[2]=a(b,c,g))return!0}}}function rb(a){return a.length>1?function(b,c,d){var e=a.length;while(e--)if(!a[e](b,c,d))return!1;return!0}:a[0]}function sb(a,b,c,d,e){for(var f,g=[],h=0,i=a.length,j=null!=b;i>h;h++)(f=a[h])&&(!c||c(f,d,e))&&(g.push(f),j&&b.push(h));return g}function tb(a,b,c,d,e,f){return d&&!d[s]&&(d=tb(d)),e&&!e[s]&&(e=tb(e,f)),fb(function(f,g,h,i){var j,k,l,m=[],n=[],o=g.length,p=f||wb(b||"*",h.nodeType?[h]:h,[]),q=!a||!f&&b?p:sb(p,m,a,h,i),r=c?e||(f?a:o||d)?[]:g:q;if(c&&c(q,r,h,i),d){j=sb(r,n),d(j,[],h,i),k=j.length;while(k--)(l=j[k])&&(r[n[k]]=!(q[n[k]]=l))}if(f){if(e||a){if(e){j=[],k=r.length;while(k--)(l=r[k])&&j.push(q[k]=l);e(null,r=[],j,i)}k=r.length;while(k--)(l=r[k])&&(j=e?I.call(f,l):m[k])>-1&&(f[j]=!(g[j]=l))}}else r=sb(r===g?r.splice(o,r.length):r),e?e(null,g,r,i):G.apply(g,r)})}function ub(a){for(var b,c,e,f=a.length,g=d.relative[a[0].type],i=g||d.relative[" "],j=g?1:0,k=qb(function(a){return a===b},i,!0),l=qb(function(a){return I.call(b,a)>-1},i,!0),m=[function(a,c,d){return!g&&(d||c!==h)||((b=c).nodeType?k(a,c,d):l(a,c,d))}];f>j;j++)if(c=d.relative[a[j].type])m=[qb(rb(m),c)];else{if(c=d.filter[a[j].type].apply(null,a[j].matches),c[s]){for(e=++j;f>e;e++)if(d.relative[a[e].type])break;return tb(j>1&&rb(m),j>1&&pb(a.slice(0,j-1).concat({value:" "===a[j-2].type?"*":""})).replace(P,"$1"),c,e>j&&ub(a.slice(j,e)),f>e&&ub(a=a.slice(e)),f>e&&pb(a))}m.push(c)}return rb(m)}function vb(a,b){var c=b.length>0,e=a.length>0,f=function(f,g,i,j,k){var m,n,o,p=0,q="0",r=f&&[],s=[],t=h,v=f||e&&d.find.TAG("*",k),w=u+=null==t?1:Math.random()||.1,x=v.length;for(k&&(h=g!==l&&g);q!==x&&null!=(m=v[q]);q++){if(e&&m){n=0;while(o=a[n++])if(o(m,g,i)){j.push(m);break}k&&(u=w)}c&&((m=!o&&m)&&p--,f&&r.push(m))}if(p+=q,c&&q!==p){n=0;while(o=b[n++])o(r,s,g,i);if(f){if(p>0)while(q--)r[q]||s[q]||(s[q]=E.call(j));s=sb(s)}G.apply(j,s),k&&!f&&s.length>0&&p+b.length>1&&db.uniqueSort(j)}return k&&(u=w,h=t),r};return c?fb(f):f}g=db.compile=function(a,b){var c,d=[],e=[],f=y[a+" "];if(!f){b||(b=ob(a)),c=b.length;while(c--)f=ub(b[c]),f[s]?d.push(f):e.push(f);f=y(a,vb(e,d))}return f};function wb(a,b,c){for(var d=0,e=b.length;e>d;d++)db(a,b[d],c);return c}function xb(a,b,e,f){var h,i,j,k,l,m=ob(a);if(!f&&1===m.length){if(i=m[0]=m[0].slice(0),i.length>2&&"ID"===(j=i[0]).type&&c.getById&&9===b.nodeType&&n&&d.relative[i[1].type]){if(b=(d.find.ID(j.matches[0].replace(ab,bb),b)||[])[0],!b)return e;a=a.slice(i.shift().value.length)}h=V.needsContext.test(a)?0:i.length;while(h--){if(j=i[h],d.relative[k=j.type])break;if((l=d.find[k])&&(f=l(j.matches[0].replace(ab,bb),$.test(i[0].type)&&mb(b.parentNode)||b))){if(i.splice(h,1),a=f.length&&pb(i),!a)return G.apply(e,f),e;break}}}return g(a,m)(f,b,!n,e,$.test(a)&&mb(b.parentNode)||b),e}return c.sortStable=s.split("").sort(z).join("")===s,c.detectDuplicates=!!j,k(),c.sortDetached=gb(function(a){return 1&a.compareDocumentPosition(l.createElement("div"))}),gb(function(a){return a.innerHTML="","#"===a.firstChild.getAttribute("href")})||hb("type|href|height|width",function(a,b,c){return c?void 0:a.getAttribute(b,"type"===b.toLowerCase()?1:2)}),c.attributes&&gb(function(a){return a.innerHTML="",a.firstChild.setAttribute("value",""),""===a.firstChild.getAttribute("value")})||hb("value",function(a,b,c){return c||"input"!==a.nodeName.toLowerCase()?void 0:a.defaultValue}),gb(function(a){return null==a.getAttribute("disabled")})||hb(J,function(a,b,c){var d;return c?void 0:a[b]===!0?b.toLowerCase():(d=a.getAttributeNode(b))&&d.specified?d.value:null}),db}(a);n.find=t,n.expr=t.selectors,n.expr[":"]=n.expr.pseudos,n.unique=t.uniqueSort,n.text=t.getText,n.isXMLDoc=t.isXML,n.contains=t.contains;var u=n.expr.match.needsContext,v=/^<(\w+)\s*\/?>(?:<\/\1>|)$/,w=/^.[^:#\[\.,]*$/;function x(a,b,c){if(n.isFunction(b))return n.grep(a,function(a,d){return!!b.call(a,d,a)!==c});if(b.nodeType)return n.grep(a,function(a){return a===b!==c});if("string"==typeof b){if(w.test(b))return n.filter(b,a,c);b=n.filter(b,a)}return n.grep(a,function(a){return n.inArray(a,b)>=0!==c})}n.filter=function(a,b,c){var d=b[0];return c&&(a=":not("+a+")"),1===b.length&&1===d.nodeType?n.find.matchesSelector(d,a)?[d]:[]:n.find.matches(a,n.grep(b,function(a){return 1===a.nodeType}))},n.fn.extend({find:function(a){var b,c=[],d=this,e=d.length;if("string"!=typeof a)return this.pushStack(n(a).filter(function(){for(b=0;e>b;b++)if(n.contains(d[b],this))return!0}));for(b=0;e>b;b++)n.find(a,d[b],c);return c=this.pushStack(e>1?n.unique(c):c),c.selector=this.selector?this.selector+" "+a:a,c},filter:function(a){return this.pushStack(x(this,a||[],!1))},not:function(a){return this.pushStack(x(this,a||[],!0))},is:function(a){return!!x(this,"string"==typeof a&&u.test(a)?n(a):a||[],!1).length}});var y,z=a.document,A=/^(?:\s*(<[\w\W]+>)[^>]*|#([\w-]*))$/,B=n.fn.init=function(a,b){var c,d;if(!a)return this;if("string"==typeof a){if(c="<"===a.charAt(0)&&">"===a.charAt(a.length-1)&&a.length>=3?[null,a,null]:A.exec(a),!c||!c[1]&&b)return!b||b.jquery?(b||y).find(a):this.constructor(b).find(a);if(c[1]){if(b=b instanceof n?b[0]:b,n.merge(this,n.parseHTML(c[1],b&&b.nodeType?b.ownerDocument||b:z,!0)),v.test(c[1])&&n.isPlainObject(b))for(c in b)n.isFunction(this[c])?this[c](b[c]):this.attr(c,b[c]);return this}if(d=z.getElementById(c[2]),d&&d.parentNode){if(d.id!==c[2])return y.find(a);this.length=1,this[0]=d}return this.context=z,this.selector=a,this}return a.nodeType?(this.context=this[0]=a,this.length=1,this):n.isFunction(a)?"undefined"!=typeof y.ready?y.ready(a):a(n):(void 0!==a.selector&&(this.selector=a.selector,this.context=a.context),n.makeArray(a,this))};B.prototype=n.fn,y=n(z);var C=/^(?:parents|prev(?:Until|All))/,D={children:!0,contents:!0,next:!0,prev:!0};n.extend({dir:function(a,b,c){var d=[],e=a[b];while(e&&9!==e.nodeType&&(void 0===c||1!==e.nodeType||!n(e).is(c)))1===e.nodeType&&d.push(e),e=e[b];return d},sibling:function(a,b){for(var c=[];a;a=a.nextSibling)1===a.nodeType&&a!==b&&c.push(a);return c}}),n.fn.extend({has:function(a){var b,c=n(a,this),d=c.length;return this.filter(function(){for(b=0;d>b;b++)if(n.contains(this,c[b]))return!0})},closest:function(a,b){for(var c,d=0,e=this.length,f=[],g=u.test(a)||"string"!=typeof a?n(a,b||this.context):0;e>d;d++)for(c=this[d];c&&c!==b;c=c.parentNode)if(c.nodeType<11&&(g?g.index(c)>-1:1===c.nodeType&&n.find.matchesSelector(c,a))){f.push(c);break}return this.pushStack(f.length>1?n.unique(f):f)},index:function(a){return a?"string"==typeof a?n.inArray(this[0],n(a)):n.inArray(a.jquery?a[0]:a,this):this[0]&&this[0].parentNode?this.first().prevAll().length:-1},add:function(a,b){return this.pushStack(n.unique(n.merge(this.get(),n(a,b))))},addBack:function(a){return this.add(null==a?this.prevObject:this.prevObject.filter(a))}});function E(a,b){do a=a[b];while(a&&1!==a.nodeType);return a}n.each({parent:function(a){var b=a.parentNode;return b&&11!==b.nodeType?b:null},parents:function(a){return n.dir(a,"parentNode")},parentsUntil:function(a,b,c){return n.dir(a,"parentNode",c)},next:function(a){return E(a,"nextSibling")},prev:function(a){return E(a,"previousSibling")},nextAll:function(a){return n.dir(a,"nextSibling")},prevAll:function(a){return n.dir(a,"previousSibling")},nextUntil:function(a,b,c){return n.dir(a,"nextSibling",c)},prevUntil:function(a,b,c){return n.dir(a,"previousSibling",c)},siblings:function(a){return n.sibling((a.parentNode||{}).firstChild,a)},children:function(a){return n.sibling(a.firstChild)},contents:function(a){return n.nodeName(a,"iframe")?a.contentDocument||a.contentWindow.document:n.merge([],a.childNodes)}},function(a,b){n.fn[a]=function(c,d){var e=n.map(this,b,c);return"Until"!==a.slice(-5)&&(d=c),d&&"string"==typeof d&&(e=n.filter(d,e)),this.length>1&&(D[a]||(e=n.unique(e)),C.test(a)&&(e=e.reverse())),this.pushStack(e)}});var F=/\S+/g,G={};function H(a){var b=G[a]={};return n.each(a.match(F)||[],function(a,c){b[c]=!0}),b}n.Callbacks=function(a){a="string"==typeof a?G[a]||H(a):n.extend({},a);var b,c,d,e,f,g,h=[],i=!a.once&&[],j=function(l){for(c=a.memory&&l,d=!0,f=g||0,g=0,e=h.length,b=!0;h&&e>f;f++)if(h[f].apply(l[0],l[1])===!1&&a.stopOnFalse){c=!1;break}b=!1,h&&(i?i.length&&j(i.shift()):c?h=[]:k.disable())},k={add:function(){if(h){var d=h.length;!function f(b){n.each(b,function(b,c){var d=n.type(c);"function"===d?a.unique&&k.has(c)||h.push(c):c&&c.length&&"string"!==d&&f(c)})}(arguments),b?e=h.length:c&&(g=d,j(c))}return this},remove:function(){return h&&n.each(arguments,function(a,c){var d;while((d=n.inArray(c,h,d))>-1)h.splice(d,1),b&&(e>=d&&e--,f>=d&&f--)}),this},has:function(a){return a?n.inArray(a,h)>-1:!(!h||!h.length)},empty:function(){return h=[],e=0,this},disable:function(){return h=i=c=void 0,this},disabled:function(){return!h},lock:function(){return i=void 0,c||k.disable(),this},locked:function(){return!i},fireWith:function(a,c){return!h||d&&!i||(c=c||[],c=[a,c.slice?c.slice():c],b?i.push(c):j(c)),this},fire:function(){return k.fireWith(this,arguments),this},fired:function(){return!!d}};return k},n.extend({Deferred:function(a){var b=[["resolve","done",n.Callbacks("once memory"),"resolved"],["reject","fail",n.Callbacks("once memory"),"rejected"],["notify","progress",n.Callbacks("memory")]],c="pending",d={state:function(){return c},always:function(){return e.done(arguments).fail(arguments),this},then:function(){var a=arguments;return n.Deferred(function(c){n.each(b,function(b,f){var g=n.isFunction(a[b])&&a[b];e[f[1]](function(){var a=g&&g.apply(this,arguments);a&&n.isFunction(a.promise)?a.promise().done(c.resolve).fail(c.reject).progress(c.notify):c[f[0]+"With"](this===d?c.promise():this,g?[a]:arguments)})}),a=null}).promise()},promise:function(a){return null!=a?n.extend(a,d):d}},e={};return d.pipe=d.then,n.each(b,function(a,f){var g=f[2],h=f[3];d[f[1]]=g.add,h&&g.add(function(){c=h},b[1^a][2].disable,b[2][2].lock),e[f[0]]=function(){return e[f[0]+"With"](this===e?d:this,arguments),this},e[f[0]+"With"]=g.fireWith}),d.promise(e),a&&a.call(e,e),e},when:function(a){var b=0,c=d.call(arguments),e=c.length,f=1!==e||a&&n.isFunction(a.promise)?e:0,g=1===f?a:n.Deferred(),h=function(a,b,c){return function(e){b[a]=this,c[a]=arguments.length>1?d.call(arguments):e,c===i?g.notifyWith(b,c):--f||g.resolveWith(b,c)}},i,j,k;if(e>1)for(i=new Array(e),j=new Array(e),k=new Array(e);e>b;b++)c[b]&&n.isFunction(c[b].promise)?c[b].promise().done(h(b,k,c)).fail(g.reject).progress(h(b,j,i)):--f;return f||g.resolveWith(k,c),g.promise()}});var I;n.fn.ready=function(a){return n.ready.promise().done(a),this},n.extend({isReady:!1,readyWait:1,holdReady:function(a){a?n.readyWait++:n.ready(!0)},ready:function(a){if(a===!0?!--n.readyWait:!n.isReady){if(!z.body)return setTimeout(n.ready);n.isReady=!0,a!==!0&&--n.readyWait>0||(I.resolveWith(z,[n]),n.fn.trigger&&n(z).trigger("ready").off("ready"))}}});function J(){z.addEventListener?(z.removeEventListener("DOMContentLoaded",K,!1),a.removeEventListener("load",K,!1)):(z.detachEvent("onreadystatechange",K),a.detachEvent("onload",K))}function K(){(z.addEventListener||"load"===event.type||"complete"===z.readyState)&&(J(),n.ready())}n.ready.promise=function(b){if(!I)if(I=n.Deferred(),"complete"===z.readyState)setTimeout(n.ready);else if(z.addEventListener)z.addEventListener("DOMContentLoaded",K,!1),a.addEventListener("load",K,!1);else{z.attachEvent("onreadystatechange",K),a.attachEvent("onload",K);var c=!1;try{c=null==a.frameElement&&z.documentElement}catch(d){}c&&c.doScroll&&!function e(){if(!n.isReady){try{c.doScroll("left")}catch(a){return setTimeout(e,50)}J(),n.ready()}}()}return I.promise(b)};var L="undefined",M;for(M in n(l))break;l.ownLast="0"!==M,l.inlineBlockNeedsLayout=!1,n(function(){var a,b,c=z.getElementsByTagName("body")[0];c&&(a=z.createElement("div"),a.style.cssText="border:0;width:0;height:0;position:absolute;top:0;left:-9999px;margin-top:1px",b=z.createElement("div"),c.appendChild(a).appendChild(b),typeof b.style.zoom!==L&&(b.style.cssText="border:0;margin:0;width:1px;padding:1px;display:inline;zoom:1",(l.inlineBlockNeedsLayout=3===b.offsetWidth)&&(c.style.zoom=1)),c.removeChild(a),a=b=null)}),function(){var a=z.createElement("div");if(null==l.deleteExpando){l.deleteExpando=!0;try{delete a.test}catch(b){l.deleteExpando=!1}}a=null}(),n.acceptData=function(a){var b=n.noData[(a.nodeName+" ").toLowerCase()],c=+a.nodeType||1;return 1!==c&&9!==c?!1:!b||b!==!0&&a.getAttribute("classid")===b};var N=/^(?:\{[\w\W]*\}|\[[\w\W]*\])$/,O=/([A-Z])/g;function P(a,b,c){if(void 0===c&&1===a.nodeType){var d="data-"+b.replace(O,"-$1").toLowerCase();if(c=a.getAttribute(d),"string"==typeof c){try{c="true"===c?!0:"false"===c?!1:"null"===c?null:+c+""===c?+c:N.test(c)?n.parseJSON(c):c}catch(e){}n.data(a,b,c)}else c=void 0}return c}function Q(a){var b;for(b in a)if(("data"!==b||!n.isEmptyObject(a[b]))&&"toJSON"!==b)return!1;return!0}function R(a,b,d,e){if(n.acceptData(a)){var f,g,h=n.expando,i=a.nodeType,j=i?n.cache:a,k=i?a[h]:a[h]&&h;if(k&&j[k]&&(e||j[k].data)||void 0!==d||"string"!=typeof b)return k||(k=i?a[h]=c.pop()||n.guid++:h),j[k]||(j[k]=i?{}:{toJSON:n.noop}),("object"==typeof b||"function"==typeof b)&&(e?j[k]=n.extend(j[k],b):j[k].data=n.extend(j[k].data,b)),g=j[k],e||(g.data||(g.data={}),g=g.data),void 0!==d&&(g[n.camelCase(b)]=d),"string"==typeof b?(f=g[b],null==f&&(f=g[n.camelCase(b)])):f=g,f -}}function S(a,b,c){if(n.acceptData(a)){var d,e,f=a.nodeType,g=f?n.cache:a,h=f?a[n.expando]:n.expando;if(g[h]){if(b&&(d=c?g[h]:g[h].data)){n.isArray(b)?b=b.concat(n.map(b,n.camelCase)):b in d?b=[b]:(b=n.camelCase(b),b=b in d?[b]:b.split(" ")),e=b.length;while(e--)delete d[b[e]];if(c?!Q(d):!n.isEmptyObject(d))return}(c||(delete g[h].data,Q(g[h])))&&(f?n.cleanData([a],!0):l.deleteExpando||g!=g.window?delete g[h]:g[h]=null)}}}n.extend({cache:{},noData:{"applet ":!0,"embed ":!0,"object ":"clsid:D27CDB6E-AE6D-11cf-96B8-444553540000"},hasData:function(a){return a=a.nodeType?n.cache[a[n.expando]]:a[n.expando],!!a&&!Q(a)},data:function(a,b,c){return R(a,b,c)},removeData:function(a,b){return S(a,b)},_data:function(a,b,c){return R(a,b,c,!0)},_removeData:function(a,b){return S(a,b,!0)}}),n.fn.extend({data:function(a,b){var c,d,e,f=this[0],g=f&&f.attributes;if(void 0===a){if(this.length&&(e=n.data(f),1===f.nodeType&&!n._data(f,"parsedAttrs"))){c=g.length;while(c--)d=g[c].name,0===d.indexOf("data-")&&(d=n.camelCase(d.slice(5)),P(f,d,e[d]));n._data(f,"parsedAttrs",!0)}return e}return"object"==typeof a?this.each(function(){n.data(this,a)}):arguments.length>1?this.each(function(){n.data(this,a,b)}):f?P(f,a,n.data(f,a)):void 0},removeData:function(a){return this.each(function(){n.removeData(this,a)})}}),n.extend({queue:function(a,b,c){var d;return a?(b=(b||"fx")+"queue",d=n._data(a,b),c&&(!d||n.isArray(c)?d=n._data(a,b,n.makeArray(c)):d.push(c)),d||[]):void 0},dequeue:function(a,b){b=b||"fx";var c=n.queue(a,b),d=c.length,e=c.shift(),f=n._queueHooks(a,b),g=function(){n.dequeue(a,b)};"inprogress"===e&&(e=c.shift(),d--),e&&("fx"===b&&c.unshift("inprogress"),delete f.stop,e.call(a,g,f)),!d&&f&&f.empty.fire()},_queueHooks:function(a,b){var c=b+"queueHooks";return n._data(a,c)||n._data(a,c,{empty:n.Callbacks("once memory").add(function(){n._removeData(a,b+"queue"),n._removeData(a,c)})})}}),n.fn.extend({queue:function(a,b){var c=2;return"string"!=typeof a&&(b=a,a="fx",c--),arguments.lengthh;h++)b(a[h],c,g?d:d.call(a[h],h,b(a[h],c)));return e?a:j?b.call(a):i?b(a[0],c):f},X=/^(?:checkbox|radio)$/i;!function(){var a=z.createDocumentFragment(),b=z.createElement("div"),c=z.createElement("input");if(b.setAttribute("className","t"),b.innerHTML="
a",l.leadingWhitespace=3===b.firstChild.nodeType,l.tbody=!b.getElementsByTagName("tbody").length,l.htmlSerialize=!!b.getElementsByTagName("link").length,l.html5Clone="<:nav>"!==z.createElement("nav").cloneNode(!0).outerHTML,c.type="checkbox",c.checked=!0,a.appendChild(c),l.appendChecked=c.checked,b.innerHTML="",l.noCloneChecked=!!b.cloneNode(!0).lastChild.defaultValue,a.appendChild(b),b.innerHTML="",l.checkClone=b.cloneNode(!0).cloneNode(!0).lastChild.checked,l.noCloneEvent=!0,b.attachEvent&&(b.attachEvent("onclick",function(){l.noCloneEvent=!1}),b.cloneNode(!0).click()),null==l.deleteExpando){l.deleteExpando=!0;try{delete b.test}catch(d){l.deleteExpando=!1}}a=b=c=null}(),function(){var b,c,d=z.createElement("div");for(b in{submit:!0,change:!0,focusin:!0})c="on"+b,(l[b+"Bubbles"]=c in a)||(d.setAttribute(c,"t"),l[b+"Bubbles"]=d.attributes[c].expando===!1);d=null}();var Y=/^(?:input|select|textarea)$/i,Z=/^key/,$=/^(?:mouse|contextmenu)|click/,_=/^(?:focusinfocus|focusoutblur)$/,ab=/^([^.]*)(?:\.(.+)|)$/;function bb(){return!0}function cb(){return!1}function db(){try{return z.activeElement}catch(a){}}n.event={global:{},add:function(a,b,c,d,e){var f,g,h,i,j,k,l,m,o,p,q,r=n._data(a);if(r){c.handler&&(i=c,c=i.handler,e=i.selector),c.guid||(c.guid=n.guid++),(g=r.events)||(g=r.events={}),(k=r.handle)||(k=r.handle=function(a){return typeof n===L||a&&n.event.triggered===a.type?void 0:n.event.dispatch.apply(k.elem,arguments)},k.elem=a),b=(b||"").match(F)||[""],h=b.length;while(h--)f=ab.exec(b[h])||[],o=q=f[1],p=(f[2]||"").split(".").sort(),o&&(j=n.event.special[o]||{},o=(e?j.delegateType:j.bindType)||o,j=n.event.special[o]||{},l=n.extend({type:o,origType:q,data:d,handler:c,guid:c.guid,selector:e,needsContext:e&&n.expr.match.needsContext.test(e),namespace:p.join(".")},i),(m=g[o])||(m=g[o]=[],m.delegateCount=0,j.setup&&j.setup.call(a,d,p,k)!==!1||(a.addEventListener?a.addEventListener(o,k,!1):a.attachEvent&&a.attachEvent("on"+o,k))),j.add&&(j.add.call(a,l),l.handler.guid||(l.handler.guid=c.guid)),e?m.splice(m.delegateCount++,0,l):m.push(l),n.event.global[o]=!0);a=null}},remove:function(a,b,c,d,e){var f,g,h,i,j,k,l,m,o,p,q,r=n.hasData(a)&&n._data(a);if(r&&(k=r.events)){b=(b||"").match(F)||[""],j=b.length;while(j--)if(h=ab.exec(b[j])||[],o=q=h[1],p=(h[2]||"").split(".").sort(),o){l=n.event.special[o]||{},o=(d?l.delegateType:l.bindType)||o,m=k[o]||[],h=h[2]&&new RegExp("(^|\\.)"+p.join("\\.(?:.*\\.|)")+"(\\.|$)"),i=f=m.length;while(f--)g=m[f],!e&&q!==g.origType||c&&c.guid!==g.guid||h&&!h.test(g.namespace)||d&&d!==g.selector&&("**"!==d||!g.selector)||(m.splice(f,1),g.selector&&m.delegateCount--,l.remove&&l.remove.call(a,g));i&&!m.length&&(l.teardown&&l.teardown.call(a,p,r.handle)!==!1||n.removeEvent(a,o,r.handle),delete k[o])}else for(o in k)n.event.remove(a,o+b[j],c,d,!0);n.isEmptyObject(k)&&(delete r.handle,n._removeData(a,"events"))}},trigger:function(b,c,d,e){var f,g,h,i,k,l,m,o=[d||z],p=j.call(b,"type")?b.type:b,q=j.call(b,"namespace")?b.namespace.split("."):[];if(h=l=d=d||z,3!==d.nodeType&&8!==d.nodeType&&!_.test(p+n.event.triggered)&&(p.indexOf(".")>=0&&(q=p.split("."),p=q.shift(),q.sort()),g=p.indexOf(":")<0&&"on"+p,b=b[n.expando]?b:new n.Event(p,"object"==typeof b&&b),b.isTrigger=e?2:3,b.namespace=q.join("."),b.namespace_re=b.namespace?new RegExp("(^|\\.)"+q.join("\\.(?:.*\\.|)")+"(\\.|$)"):null,b.result=void 0,b.target||(b.target=d),c=null==c?[b]:n.makeArray(c,[b]),k=n.event.special[p]||{},e||!k.trigger||k.trigger.apply(d,c)!==!1)){if(!e&&!k.noBubble&&!n.isWindow(d)){for(i=k.delegateType||p,_.test(i+p)||(h=h.parentNode);h;h=h.parentNode)o.push(h),l=h;l===(d.ownerDocument||z)&&o.push(l.defaultView||l.parentWindow||a)}m=0;while((h=o[m++])&&!b.isPropagationStopped())b.type=m>1?i:k.bindType||p,f=(n._data(h,"events")||{})[b.type]&&n._data(h,"handle"),f&&f.apply(h,c),f=g&&h[g],f&&f.apply&&n.acceptData(h)&&(b.result=f.apply(h,c),b.result===!1&&b.preventDefault());if(b.type=p,!e&&!b.isDefaultPrevented()&&(!k._default||k._default.apply(o.pop(),c)===!1)&&n.acceptData(d)&&g&&d[p]&&!n.isWindow(d)){l=d[g],l&&(d[g]=null),n.event.triggered=p;try{d[p]()}catch(r){}n.event.triggered=void 0,l&&(d[g]=l)}return b.result}},dispatch:function(a){a=n.event.fix(a);var b,c,e,f,g,h=[],i=d.call(arguments),j=(n._data(this,"events")||{})[a.type]||[],k=n.event.special[a.type]||{};if(i[0]=a,a.delegateTarget=this,!k.preDispatch||k.preDispatch.call(this,a)!==!1){h=n.event.handlers.call(this,a,j),b=0;while((f=h[b++])&&!a.isPropagationStopped()){a.currentTarget=f.elem,g=0;while((e=f.handlers[g++])&&!a.isImmediatePropagationStopped())(!a.namespace_re||a.namespace_re.test(e.namespace))&&(a.handleObj=e,a.data=e.data,c=((n.event.special[e.origType]||{}).handle||e.handler).apply(f.elem,i),void 0!==c&&(a.result=c)===!1&&(a.preventDefault(),a.stopPropagation()))}return k.postDispatch&&k.postDispatch.call(this,a),a.result}},handlers:function(a,b){var c,d,e,f,g=[],h=b.delegateCount,i=a.target;if(h&&i.nodeType&&(!a.button||"click"!==a.type))for(;i!=this;i=i.parentNode||this)if(1===i.nodeType&&(i.disabled!==!0||"click"!==a.type)){for(e=[],f=0;h>f;f++)d=b[f],c=d.selector+" ",void 0===e[c]&&(e[c]=d.needsContext?n(c,this).index(i)>=0:n.find(c,this,null,[i]).length),e[c]&&e.push(d);e.length&&g.push({elem:i,handlers:e})}return h]","i"),ib=/^\s+/,jb=/<(?!area|br|col|embed|hr|img|input|link|meta|param)(([\w:]+)[^>]*)\/>/gi,kb=/<([\w:]+)/,lb=/\s*$/g,sb={option:[1,""],legend:[1,"
","
"],area:[1,"",""],param:[1,"",""],thead:[1,"","
"],tr:[2,"","
"],col:[2,"","
"],td:[3,"","
"],_default:l.htmlSerialize?[0,"",""]:[1,"X
","
"]},tb=eb(z),ub=tb.appendChild(z.createElement("div"));sb.optgroup=sb.option,sb.tbody=sb.tfoot=sb.colgroup=sb.caption=sb.thead,sb.th=sb.td;function vb(a,b){var c,d,e=0,f=typeof a.getElementsByTagName!==L?a.getElementsByTagName(b||"*"):typeof a.querySelectorAll!==L?a.querySelectorAll(b||"*"):void 0;if(!f)for(f=[],c=a.childNodes||a;null!=(d=c[e]);e++)!b||n.nodeName(d,b)?f.push(d):n.merge(f,vb(d,b));return void 0===b||b&&n.nodeName(a,b)?n.merge([a],f):f}function wb(a){X.test(a.type)&&(a.defaultChecked=a.checked)}function xb(a,b){return n.nodeName(a,"table")&&n.nodeName(11!==b.nodeType?b:b.firstChild,"tr")?a.getElementsByTagName("tbody")[0]||a.appendChild(a.ownerDocument.createElement("tbody")):a}function yb(a){return a.type=(null!==n.find.attr(a,"type"))+"/"+a.type,a}function zb(a){var b=qb.exec(a.type);return b?a.type=b[1]:a.removeAttribute("type"),a}function Ab(a,b){for(var c,d=0;null!=(c=a[d]);d++)n._data(c,"globalEval",!b||n._data(b[d],"globalEval"))}function Bb(a,b){if(1===b.nodeType&&n.hasData(a)){var c,d,e,f=n._data(a),g=n._data(b,f),h=f.events;if(h){delete g.handle,g.events={};for(c in h)for(d=0,e=h[c].length;e>d;d++)n.event.add(b,c,h[c][d])}g.data&&(g.data=n.extend({},g.data))}}function Cb(a,b){var c,d,e;if(1===b.nodeType){if(c=b.nodeName.toLowerCase(),!l.noCloneEvent&&b[n.expando]){e=n._data(b);for(d in e.events)n.removeEvent(b,d,e.handle);b.removeAttribute(n.expando)}"script"===c&&b.text!==a.text?(yb(b).text=a.text,zb(b)):"object"===c?(b.parentNode&&(b.outerHTML=a.outerHTML),l.html5Clone&&a.innerHTML&&!n.trim(b.innerHTML)&&(b.innerHTML=a.innerHTML)):"input"===c&&X.test(a.type)?(b.defaultChecked=b.checked=a.checked,b.value!==a.value&&(b.value=a.value)):"option"===c?b.defaultSelected=b.selected=a.defaultSelected:("input"===c||"textarea"===c)&&(b.defaultValue=a.defaultValue)}}n.extend({clone:function(a,b,c){var d,e,f,g,h,i=n.contains(a.ownerDocument,a);if(l.html5Clone||n.isXMLDoc(a)||!hb.test("<"+a.nodeName+">")?f=a.cloneNode(!0):(ub.innerHTML=a.outerHTML,ub.removeChild(f=ub.firstChild)),!(l.noCloneEvent&&l.noCloneChecked||1!==a.nodeType&&11!==a.nodeType||n.isXMLDoc(a)))for(d=vb(f),h=vb(a),g=0;null!=(e=h[g]);++g)d[g]&&Cb(e,d[g]);if(b)if(c)for(h=h||vb(a),d=d||vb(f),g=0;null!=(e=h[g]);g++)Bb(e,d[g]);else Bb(a,f);return d=vb(f,"script"),d.length>0&&Ab(d,!i&&vb(a,"script")),d=h=e=null,f},buildFragment:function(a,b,c,d){for(var e,f,g,h,i,j,k,m=a.length,o=eb(b),p=[],q=0;m>q;q++)if(f=a[q],f||0===f)if("object"===n.type(f))n.merge(p,f.nodeType?[f]:f);else if(mb.test(f)){h=h||o.appendChild(b.createElement("div")),i=(kb.exec(f)||["",""])[1].toLowerCase(),k=sb[i]||sb._default,h.innerHTML=k[1]+f.replace(jb,"<$1>")+k[2],e=k[0];while(e--)h=h.lastChild;if(!l.leadingWhitespace&&ib.test(f)&&p.push(b.createTextNode(ib.exec(f)[0])),!l.tbody){f="table"!==i||lb.test(f)?""!==k[1]||lb.test(f)?0:h:h.firstChild,e=f&&f.childNodes.length;while(e--)n.nodeName(j=f.childNodes[e],"tbody")&&!j.childNodes.length&&f.removeChild(j)}n.merge(p,h.childNodes),h.textContent="";while(h.firstChild)h.removeChild(h.firstChild);h=o.lastChild}else p.push(b.createTextNode(f));h&&o.removeChild(h),l.appendChecked||n.grep(vb(p,"input"),wb),q=0;while(f=p[q++])if((!d||-1===n.inArray(f,d))&&(g=n.contains(f.ownerDocument,f),h=vb(o.appendChild(f),"script"),g&&Ab(h),c)){e=0;while(f=h[e++])pb.test(f.type||"")&&c.push(f)}return h=null,o},cleanData:function(a,b){for(var d,e,f,g,h=0,i=n.expando,j=n.cache,k=l.deleteExpando,m=n.event.special;null!=(d=a[h]);h++)if((b||n.acceptData(d))&&(f=d[i],g=f&&j[f])){if(g.events)for(e in g.events)m[e]?n.event.remove(d,e):n.removeEvent(d,e,g.handle);j[f]&&(delete j[f],k?delete d[i]:typeof d.removeAttribute!==L?d.removeAttribute(i):d[i]=null,c.push(f))}}}),n.fn.extend({text:function(a){return W(this,function(a){return void 0===a?n.text(this):this.empty().append((this[0]&&this[0].ownerDocument||z).createTextNode(a))},null,a,arguments.length)},append:function(){return this.domManip(arguments,function(a){if(1===this.nodeType||11===this.nodeType||9===this.nodeType){var b=xb(this,a);b.appendChild(a)}})},prepend:function(){return this.domManip(arguments,function(a){if(1===this.nodeType||11===this.nodeType||9===this.nodeType){var b=xb(this,a);b.insertBefore(a,b.firstChild)}})},before:function(){return this.domManip(arguments,function(a){this.parentNode&&this.parentNode.insertBefore(a,this)})},after:function(){return this.domManip(arguments,function(a){this.parentNode&&this.parentNode.insertBefore(a,this.nextSibling)})},remove:function(a,b){for(var c,d=a?n.filter(a,this):this,e=0;null!=(c=d[e]);e++)b||1!==c.nodeType||n.cleanData(vb(c)),c.parentNode&&(b&&n.contains(c.ownerDocument,c)&&Ab(vb(c,"script")),c.parentNode.removeChild(c));return this},empty:function(){for(var a,b=0;null!=(a=this[b]);b++){1===a.nodeType&&n.cleanData(vb(a,!1));while(a.firstChild)a.removeChild(a.firstChild);a.options&&n.nodeName(a,"select")&&(a.options.length=0)}return this},clone:function(a,b){return a=null==a?!1:a,b=null==b?a:b,this.map(function(){return n.clone(this,a,b)})},html:function(a){return W(this,function(a){var b=this[0]||{},c=0,d=this.length;if(void 0===a)return 1===b.nodeType?b.innerHTML.replace(gb,""):void 0;if(!("string"!=typeof a||nb.test(a)||!l.htmlSerialize&&hb.test(a)||!l.leadingWhitespace&&ib.test(a)||sb[(kb.exec(a)||["",""])[1].toLowerCase()])){a=a.replace(jb,"<$1>");try{for(;d>c;c++)b=this[c]||{},1===b.nodeType&&(n.cleanData(vb(b,!1)),b.innerHTML=a);b=0}catch(e){}}b&&this.empty().append(a)},null,a,arguments.length)},replaceWith:function(){var a=arguments[0];return this.domManip(arguments,function(b){a=this.parentNode,n.cleanData(vb(this)),a&&a.replaceChild(b,this)}),a&&(a.length||a.nodeType)?this:this.remove()},detach:function(a){return this.remove(a,!0)},domManip:function(a,b){a=e.apply([],a);var c,d,f,g,h,i,j=0,k=this.length,m=this,o=k-1,p=a[0],q=n.isFunction(p);if(q||k>1&&"string"==typeof p&&!l.checkClone&&ob.test(p))return this.each(function(c){var d=m.eq(c);q&&(a[0]=p.call(this,c,d.html())),d.domManip(a,b)});if(k&&(i=n.buildFragment(a,this[0].ownerDocument,!1,this),c=i.firstChild,1===i.childNodes.length&&(i=c),c)){for(g=n.map(vb(i,"script"),yb),f=g.length;k>j;j++)d=i,j!==o&&(d=n.clone(d,!0,!0),f&&n.merge(g,vb(d,"script"))),b.call(this[j],d,j);if(f)for(h=g[g.length-1].ownerDocument,n.map(g,zb),j=0;f>j;j++)d=g[j],pb.test(d.type||"")&&!n._data(d,"globalEval")&&n.contains(h,d)&&(d.src?n._evalUrl&&n._evalUrl(d.src):n.globalEval((d.text||d.textContent||d.innerHTML||"").replace(rb,"")));i=c=null}return this}}),n.each({appendTo:"append",prependTo:"prepend",insertBefore:"before",insertAfter:"after",replaceAll:"replaceWith"},function(a,b){n.fn[a]=function(a){for(var c,d=0,e=[],g=n(a),h=g.length-1;h>=d;d++)c=d===h?this:this.clone(!0),n(g[d])[b](c),f.apply(e,c.get());return this.pushStack(e)}});var Db,Eb={};function Fb(b,c){var d=n(c.createElement(b)).appendTo(c.body),e=a.getDefaultComputedStyle?a.getDefaultComputedStyle(d[0]).display:n.css(d[0],"display");return d.detach(),e}function Gb(a){var b=z,c=Eb[a];return c||(c=Fb(a,b),"none"!==c&&c||(Db=(Db||n(" + + {{else}} + + + {{end}} + + + + + {{ partial "footer.html" . }} diff --git a/themes/apachecon/layouts/partials/countdown.html b/themes/apachecon/layouts/partials/countdown.html new file mode 100644 index 000000000..ceddd772f --- /dev/null +++ b/themes/apachecon/layouts/partials/countdown.html @@ -0,0 +1,57 @@ +
+ {{ $array := slice "weeks" "days" "hours" "minutes" "seconds" }} + {{ range $index, $element := $array }} +
+
+ + + + +
00
+
+
{{ i18n $element }}
+
+ {{ end }} +
+ + \ No newline at end of file diff --git a/themes/apachecon/layouts/partials/footer.html b/themes/apachecon/layouts/partials/footer.html index a4c2ce847..cfb28114f 100644 --- a/themes/apachecon/layouts/partials/footer.html +++ b/themes/apachecon/layouts/partials/footer.html @@ -19,13 +19,20 @@

Connect with us

-
-
- -

ApacheCon operates under the terms of the Apache Software Foundation Code of Conduct.

-
-

Copyright 2021, The Apache Software Foundation, Licensed under the Apache License, Version 2.0.
Apache and the Apache feather logo are trademarks of The Apache Software Foundation.

+
+ + + + + + + + + +
+

Community Over Code operates under the terms of the Apache Software Foundation Code of Conduct.

+

Copyright 2026, The Apache Software Foundation, Licensed under the Apache License, Version 2.0.
Apache and the Apache feather logo are trademarks of The Apache Software Foundation. - \ No newline at end of file + diff --git a/themes/apachecon/layouts/partials/header.html b/themes/apachecon/layouts/partials/header.html index f6e20cfc6..737257ad3 100644 --- a/themes/apachecon/layouts/partials/header.html +++ b/themes/apachecon/layouts/partials/header.html @@ -1,34 +1,32 @@ - - - - ApacheCon @Home + {{if .IsHome}}{{ $.Site.Title }}{{else}}{{.Title}} - {{ $.Site.Title }}{{end}} - Powered by {{$.Site.Params.author}} - - + + + + - - + + + - + + + + + + {{ $style := resources.Get "scss/main.scss" | toCSS | minify | fingerprint }} + + {{ $acasia2021 := resources.Get "scss/acasia2021.scss" | toCSS | minify | fingerprint }} + + {{ $schedule := resources.Get "scss/schedule.scss" | toCSS | minify | fingerprint }} + + + - - - - - \ No newline at end of file + + diff --git a/themes/apachecon/layouts/partials/nav.html b/themes/apachecon/layouts/partials/nav.html index 4b8f75b01..d4ed3d9db 100644 --- a/themes/apachecon/layouts/partials/nav.html +++ b/themes/apachecon/layouts/partials/nav.html @@ -1,10 +1,71 @@ -