Bake provenance JSON into images and auto-populate post_results.py args

.github/actions/apply-provenance-labels/action.yml

@@ -0,0 +1,61 @@
+name: Apply provenance labels
+description: >-
+  Re-tag an already-built image through provenance_labels.dockerfile so it
+  carries the velox-testing provenance labels and baked /opt/velox-testing/provenance.json.
+
+inputs:
+  base_image:
+    description: Fully-qualified reference of the already-built image to layer provenance onto.
+    required: true
+  tags:
+    description: Tag(s) to push the provenance-labelled image under (newline-separated).
+    required: true
+  presto_sha:
+    description: Presto commit SHA.
+    required: true
+  presto_repository:
+    description: Presto repository.
+    required: false
+    default: ""
+  presto_branch:
+    description: Presto branch (empty on CI, which knows only the SHA).
+    required: false
+    default: ""
+  velox_sha:
+    description: Velox commit SHA (empty for coordinator images, which do not build Velox).
+    required: false
+    default: ""
+  velox_repository:
+    description: Velox repository.
+    required: false
+    default: ""
+  velox_branch:
+    description: Velox branch (empty on CI).
+    required: false
+    default: ""
+
+outputs:
+  digest:
+    description: Digest of the pushed provenance-labelled image.
+    value: ${{ steps.push.outputs.digest }}
+
+runs:
+  using: composite
+  steps:
+    - name: Build and push provenance-labelled image
+      id: push
+      uses: docker/build-push-action@263435318d21b8e681c14492fe198d362a7d2c83  # v6.18.0
+      with:
+        context: velox-testing/presto/docker
+        file: velox-testing/presto/docker/provenance_labels.dockerfile
+        build-args: |
+          BASE_IMAGE=${{ inputs.base_image }}
+          PRESTO_SHA=${{ inputs.presto_sha }}
+          PRESTO_BRANCH=${{ inputs.presto_branch }}
+          PRESTO_REPOSITORY=${{ inputs.presto_repository }}
+          VELOX_SHA=${{ inputs.velox_sha }}
+          VELOX_BRANCH=${{ inputs.velox_branch }}
+          VELOX_REPOSITORY=${{ inputs.velox_repository }}
+        outputs: type=registry,compression=zstd,force-compression=true,compression-level=15
+        push: true
+        tags: ${{ inputs.tags }}

.github/workflows/presto-build.yml

@@ -118,28 +118,31 @@ jobs:
           cp -r velox/CMake presto-deps-context/velox/CMake
 
       - name: Build and push Presto deps image
-        id: push
         uses: docker/build-push-action@263435318d21b8e681c14492fe198d362a7d2c83  # v6.18.0
         with:
           context: presto-deps-context
           file: presto/presto-native-execution/scripts/dockerfiles/centos-dependency.dockerfile
           build-args: |
             CUDA_VERSION=${{ matrix.cuda_version }}
             ARM_BUILD_TARGET=generic
-          labels: |
-            velox-testing.presto.sha=${{ inputs.presto_sha }}
-            velox-testing.presto.branch=
-            velox-testing.presto.repository=${{ inputs.presto_repository }}
-            velox-testing.velox.sha=${{ inputs.velox_sha }}
-            velox-testing.velox.branch=
-            velox-testing.velox.repository=${{ inputs.velox_repository }}
           cache-from: type=registry,ref=${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}:cache-presto-deps-cuda${{ matrix.cuda_version }}-${{ matrix.arch }}
           cache-to: type=registry,ref=${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}:cache-presto-deps-cuda${{ matrix.cuda_version }}-${{ matrix.arch }},mode=max
           outputs: type=registry,compression=zstd,force-compression=true,compression-level=15
           push: true
           tags: |
             ${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}:presto-deps-${{ inputs.presto_short_sha }}-velox-${{ inputs.velox_short_sha }}-cuda${{ matrix.cuda_version }}-${{ inputs.date }}-${{ inputs.build_variant }}-${{ github.run_id }}-${{ matrix.arch }}
 
+      - name: Apply provenance labels to Presto deps image
+        id: push
+        uses: ./velox-testing/.github/actions/apply-provenance-labels
+        with:
+          base_image: ${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}:presto-deps-${{ inputs.presto_short_sha }}-velox-${{ inputs.velox_short_sha }}-cuda${{ matrix.cuda_version }}-${{ inputs.date }}-${{ inputs.build_variant }}-${{ github.run_id }}-${{ matrix.arch }}
+          tags: ${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}:presto-deps-${{ inputs.presto_short_sha }}-velox-${{ inputs.velox_short_sha }}-cuda${{ matrix.cuda_version }}-${{ inputs.date }}-${{ inputs.build_variant }}-${{ github.run_id }}-${{ matrix.arch }}
+          presto_sha: ${{ inputs.presto_sha }}
+          presto_repository: ${{ inputs.presto_repository }}
+          velox_sha: ${{ inputs.velox_sha }}
+          velox_repository: ${{ inputs.velox_repository }}
+
       - name: Generate artifact attestation
         uses: actions/attest-build-provenance@96278af6caaf10aea03fd8d33a09a777ca52d62f  # v3.2.0
         with:
@@ -187,25 +190,26 @@ jobs:
           echo "tag=presto-coordinator-${PRESTO_SHORT_SHA}-${BUILD_DATE}-${BUILD_VARIANT}-${GITHUB_RUN_ID}-${ARCH}" >> "$GITHUB_OUTPUT"
 
       - name: Build and push Presto coordinator image
-        id: push
         uses: docker/build-push-action@263435318d21b8e681c14492fe198d362a7d2c83  # v6.18.0
         with:
           context: presto/docker
           file: presto/docker/Dockerfile
           build-args: |
             PRESTO_VERSION=${{ inputs.presto_short_sha }}
-          labels: |
-            velox-testing.presto.sha=${{ inputs.presto_sha }}
-            velox-testing.presto.branch=
-            velox-testing.presto.repository=${{ inputs.presto_repository }}
-            velox-testing.velox.sha=
-            velox-testing.velox.branch=
-            velox-testing.velox.repository=
           outputs: type=registry,compression=zstd,force-compression=true,compression-level=15
           push: true
           tags: |
             ${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}:${{ steps.tag.outputs.tag }}
 
+      - name: Apply provenance labels to Presto coordinator image
+        id: push
+        uses: ./velox-testing/.github/actions/apply-provenance-labels
+        with:
+          base_image: ${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}:${{ steps.tag.outputs.tag }}
+          tags: ${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}:${{ steps.tag.outputs.tag }}
+          presto_sha: ${{ inputs.presto_sha }}
+          presto_repository: ${{ inputs.presto_repository }}
+
       - name: Generate artifact attestation
         uses: actions/attest-build-provenance@96278af6caaf10aea03fd8d33a09a777ca52d62f  # v3.2.0
         with:

benchmark_reporting_tools/post_results.py

@@ -92,6 +92,12 @@ class BenchmarkMetadata:
     num_drivers: int | None = None
     gpu_name: str | None = None
     image_digest: str | None = None
+    presto_sha: str | None = None
+    presto_branch: str | None = None
+    presto_repo: str | None = None
+    velox_sha: str | None = None
+    velox_branch: str | None = None
+    velox_repo: str | None = None
 
     @classmethod
     def from_parsed(cls, raw: dict) -> "BenchmarkMetadata":
@@ -298,22 +304,22 @@ def _parse_args() -> argparse.Namespace:
     parser.add_argument(
         "--velox-branch",
         default=None,
-        help="Velox branch used to build the worker image.",
+        help="Velox branch used to build the worker image. Defaults to value from image labels in benchmark_result.json.",
     )
     parser.add_argument(
         "--velox-repo",
         default=None,
-        help="Velox repository used to build the worker image.",
+        help="Velox repository used to build the worker image. Defaults to value from image labels in benchmark_result.json.",
     )
     parser.add_argument(
         "--presto-branch",
         default=None,
-        help="Presto branch used to build the worker image.",
+        help="Presto branch used to build the worker image. Defaults to value from image labels in benchmark_result.json.",
     )
     parser.add_argument(
         "--presto-repo",
         default=None,
-        help="Presto repository used to build the worker image.",
+        help="Presto repository used to build the worker image. Defaults to value from image labels in benchmark_result.json.",
     )
     parser.add_argument(
         "--concurrency-streams",
@@ -387,8 +393,10 @@ def _build_submission_payload(
     validation_results: dict | None = None,
     velox_branch: str | None = None,
     velox_repo: str | None = None,
+    velox_sha: str | None = None,
     presto_branch: str | None = None,
     presto_repo: str | None = None,
+    presto_sha: str | None = None,
     labels: list[str] | None = None,
     notes: str | None = None,
 ) -> dict:
@@ -413,8 +421,15 @@ def _build_submission_payload(
     # Use placeholders for version info if not provided
     if version is None:
         version = "unknown"
+    # Derive commit_hash from the image provenance SHAs when not explicitly given:
+    # "presto-<sha>-velox-<sha>" (worker), "presto-<sha>" (coordinator, no velox).
     if commit_hash is None:
-        commit_hash = "unknown"
+        commit_parts = []
+        if presto_sha:
+            commit_parts.append(f"presto-{presto_sha}")
+        if velox_sha:
+            commit_parts.append(f"velox-{velox_sha}")
+        commit_hash = "-".join(commit_parts) if commit_parts else "unknown"
 
     # Build query logs from results
     query_logs = []
@@ -504,14 +519,19 @@ def _get_validation_result(query_name):
     }
 
     engine_config_payload = engine_config.serialize() if engine_config else {}
-    if velox_branch or velox_repo or presto_branch or presto_repo:
-        engine_config_payload = {
-            **engine_config_payload,
-            "velox_branch": velox_branch,
-            "velox_repo": velox_repo,
-            "presto_branch": presto_branch,
-            "presto_repo": presto_repo,
-        }
+    # Add image provenance as separate entries; omit empties so a coordinator-only
+    # image (no velox fields) contributes just its presto entries.
+    provenance_fields = {
+        "presto_branch": presto_branch,
+        "presto_sha": presto_sha,
+        "presto_repo": presto_repo,
+        "velox_branch": velox_branch,
+        "velox_sha": velox_sha,
+        "velox_repo": velox_repo,
+    }
+    provenance_fields = {k: v for k, v in provenance_fields.items() if v}
+    if provenance_fields:
+        engine_config_payload = {**engine_config_payload, **provenance_fields}
 
     payload: dict = {
         "sku_name": sku_name,
@@ -746,6 +766,15 @@ async def _process_benchmark_dir(
         )
         return 1
 
+    # Fall back to image provenance labels captured in the benchmark context.
+    velox_branch = velox_branch or benchmark_metadata.velox_branch
+    velox_repo = velox_repo or benchmark_metadata.velox_repo
+    presto_branch = presto_branch or benchmark_metadata.presto_branch
+    presto_repo = presto_repo or benchmark_metadata.presto_repo
+    # SHAs come only from the baked image provenance (no CLI override).
+    velox_sha = benchmark_metadata.velox_sha
+    presto_sha = benchmark_metadata.presto_sha
+
     # Resolve config directory: explicit override → auto-detect from variant
     effective_config_dir = config_dir
     variant = _ENGINE_TO_VARIANT.get(benchmark_metadata.engine)
@@ -849,8 +878,10 @@ async def _process_benchmark_dir(
                 validation_results=validation_results,
                 velox_branch=velox_branch,
                 velox_repo=velox_repo,
+                velox_sha=velox_sha,
                 presto_branch=presto_branch,
                 presto_repo=presto_repo,
+                presto_sha=presto_sha,
                 labels=labels,
                 notes=notes,
             )

common/testing/performance_benchmarks/benchmark_keys.py

@@ -30,3 +30,9 @@ class BenchmarkKeys(str, Enum):
     KIND_KEY = "kind"
     GPU_COUNT_KEY = "gpu_count"
     DATASET_NAME_KEY = "dataset_name"
+    PRESTO_SHA_KEY = "presto_sha"
+    PRESTO_BRANCH_KEY = "presto_branch"
+    PRESTO_REPO_KEY = "presto_repo"
+    VELOX_SHA_KEY = "velox_sha"
+    VELOX_BRANCH_KEY = "velox_branch"
+    VELOX_REPO_KEY = "velox_repo"

presto/docker/launch_coordinator.sh

@@ -7,6 +7,13 @@ set -e
 LOGS_DIR="/opt/presto-server/logs"
 mkdir -p "${LOGS_DIR}"
 : "${SERVER_START_TIMESTAMP:?SERVER_START_TIMESTAMP must be set before starting the container}"
+
+# Surface baked-in image provenance in the shared logs dir so the host-side
+# pytest (Docker) or in-container pytest (SLURM) can read it via LOGS_DIR.
+if [ -f /opt/velox-testing/provenance.json ]; then
+  cp /opt/velox-testing/provenance.json "${LOGS_DIR}/coordinator_provenance.json"
+fi
+
 log_file="${LOGS_DIR}/coordinator_${SERVER_START_TIMESTAMP}.log"
 
 exec /opt/presto-server/bin/launcher run >> "${log_file}" 2>&1

presto/docker/launch_presto_servers.sh

@@ -10,6 +10,12 @@ LOGS_DIR="/opt/presto-server/logs"
 mkdir -p "${LOGS_DIR}"
 : "${SERVER_START_TIMESTAMP:?SERVER_START_TIMESTAMP must be set before starting the container}"
 
+# Surface baked-in image provenance in the shared logs dir so the host-side
+# pytest (Docker) or in-container pytest (SLURM) can read it via LOGS_DIR.
+if [ -f /opt/velox-testing/provenance.json ]; then
+  cp /opt/velox-testing/provenance.json "${LOGS_DIR}/worker_provenance.json"
+fi
+
 ETC_BASE="/opt/presto-server/etc"
 
 # Resolve the NUMA node for a worker and launch presto_server pinned to it.

presto/docker/native_build.dockerfile

@@ -124,5 +124,10 @@ LABEL velox-testing.presto.sha=${PRESTO_SHA} \
       velox-testing.velox.sha=${VELOX_SHA} \
       velox-testing.velox.branch=${VELOX_BRANCH} \
       velox-testing.velox.repository=${VELOX_REPOSITORY}
+RUN mkdir -p /opt/velox-testing && \
+    printf '{"presto_sha":"%s","presto_branch":"%s","presto_repo":"%s","velox_sha":"%s","velox_branch":"%s","velox_repo":"%s"}\n' \
+    "${PRESTO_SHA}" "${PRESTO_BRANCH}" "${PRESTO_REPOSITORY}" \
+    "${VELOX_SHA}" "${VELOX_BRANCH}" "${VELOX_REPOSITORY}" \
+    > /opt/velox-testing/provenance.json
 
 CMD ["bash", "/opt/presto_profiling_wrapper.sh"]

presto/docker/provenance_labels.dockerfile

@@ -1,4 +1,4 @@
-ARG BASE_IMAGE
+ARG BASE_IMAGE=invalid
 FROM ${BASE_IMAGE}
 
 ARG PRESTO_SHA
@@ -13,3 +13,8 @@ LABEL velox-testing.presto.sha=${PRESTO_SHA} \
       velox-testing.velox.sha=${VELOX_SHA} \
       velox-testing.velox.branch=${VELOX_BRANCH} \
       velox-testing.velox.repository=${VELOX_REPOSITORY}
+RUN mkdir -p /opt/velox-testing && \
+    printf '{"presto_sha":"%s","presto_branch":"%s","presto_repo":"%s","velox_sha":"%s","velox_branch":"%s","velox_repo":"%s"}\n' \
+    "${PRESTO_SHA}" "${PRESTO_BRANCH}" "${PRESTO_REPOSITORY}" \
+    "${VELOX_SHA}" "${VELOX_BRANCH}" "${VELOX_REPOSITORY}" \
+    > /opt/velox-testing/provenance.json

presto/slurm/presto-nvl72/functions.sh

@@ -135,7 +135,8 @@ ${CONFIGS}/etc_coordinator/node.properties:/opt/presto-server/etc/node.propertie
 ${CONFIGS}/etc_coordinator/config_native.properties:/opt/presto-server/etc/config.properties,\
 ${CONFIGS}/etc_coordinator/catalog/hive.properties:/opt/presto-server/etc/catalog/hive.properties,\
 ${DATA}:/var/lib/presto/data/hive/data/user_data,\
-${VT_ROOT}/.hive_metastore:/var/lib/presto/data/hive/metastore${extra_mounts} \
+${VT_ROOT}/.hive_metastore:/var/lib/presto/data/hive/metastore,\
+${LOGS}:/opt/presto-server/logs${extra_mounts} \
 -- bash -lc "unset JAVA_HOME; export JAVA_HOME=/usr/lib/jvm/jre-17-openjdk; export PATH=/usr/lib/jvm/jre-17-openjdk/bin:\$PATH; ${script}" >> ${LOGS}/${log_file} 2>&1 &
     else
         srun -w $COORD --ntasks=1 --overlap \
@@ -149,8 +150,9 @@ ${CONFIGS}/etc_coordinator/node.properties:/opt/presto-server/etc/node.propertie
 ${CONFIGS}/etc_coordinator/config_native.properties:/opt/presto-server/etc/config.properties,\
 ${CONFIGS}/etc_coordinator/catalog/hive.properties:/opt/presto-server/etc/catalog/hive.properties,\
 ${DATA}:/var/lib/presto/data/hive/data/user_data,\
+${VT_ROOT}/.hive_metastore:/var/lib/presto/data/hive/metastore,\
 ${LOGS}:/var/log/nsys,\
-${VT_ROOT}/.hive_metastore:/var/lib/presto/data/hive/metastore${extra_mounts} \
+${LOGS}:/opt/presto-server/logs${extra_mounts} \
 -- bash -lc "unset JAVA_HOME; export JAVA_HOME=/usr/lib/jvm/jre-17-openjdk; export PATH=/usr/lib/jvm/jre-17-openjdk/bin:\$PATH; ${script}" >> ${LOGS}/${log_file} 2>&1
     fi
 }
@@ -173,6 +175,13 @@ unset CONFIG NODE_CONFIG PRESTO_ETC JAVA_TOOL_OPTIONS JDK_JAVA_OPTIONS _JAVA_OPT
 
 export JAVA_HOME=/usr
 export PATH=/usr/bin:$PATH
+
+# Surface baked-in image provenance into the shared logs dir for the in-container
+# pytest in run_benchmark.sh to read via LOGS_DIR.
+if [ -f /opt/velox-testing/provenance.json ]; then
+  cp /opt/velox-testing/provenance.json "/opt/presto-server/logs/coordinator_provenance.json"
+fi
+
 /opt/presto-server/bin/launcher run & srv=$!
 
 # wait for JVM to appear
@@ -345,6 +354,7 @@ ${worker_hive}:/opt/presto-server/etc/catalog/hive.properties,\
 ${worker_data}:/var/lib/presto/data,\
 ${DATA}:/var/lib/presto/data/hive/data/user_data,\
 ${VT_ROOT}/.hive_metastore:/var/lib/presto/data/hive/metastore,\
+${LOGS}:/opt/presto-server/logs,\
 ${WORKER_ENV_FILE}:${vt_worker_env_file},\
 ${LOGS}:${vt_cufile_log_dir},\
 ${LOGS}:${vt_nsys_report_dir}${driver_mounts}${gds_mounts:+,${gds_mounts}}${worker_extra_mounts} \
@@ -373,6 +383,10 @@ echo \"Worker ${worker_id}: CUFILE_LOGFILE_PATH=\${CUFILE_LOGFILE_PATH:-unset}\"
 echo \"Worker ${worker_id}: KVIKIO_TASK_SIZE=\${KVIKIO_TASK_SIZE:-unset}\"
 echo \"Worker ${worker_id}: KVIKIO_NTHREADS=\${KVIKIO_NTHREADS:-unset}\"
 
+if [ -f /opt/velox-testing/provenance.json ]; then
+    cp /opt/velox-testing/provenance.json /opt/presto-server/logs/worker_provenance.json
+fi
+
 if [[ -n '${nsys_bin}' ]]; then
     (
         echo \"Worker ${worker_id}: nsys subshell started\"
@@ -573,6 +587,7 @@ function run_queries {
     export PRESTO_DATA_DIR=/var/lib/presto/data/hive/data/user_data; \
     export MINIFORGE_HOME=/workspace/miniforge3; \
     export HOME=/workspace; \
+    export LOGS_DIR=/opt/presto-server/logs; \
     cd /workspace/presto/scripts; \
     ./run_benchmark.sh -b tpch -s tpchsf${scale_factor} -i ${num_iterations} ${extra_args[*]} \
         --hostname ${COORD} --port $PORT -o ${container_script_dir}/result_dir --skip-drop-cache" "cli"