docs: add example Grafana dashboard for GenAI metrics - #2374
Open
baiyutang wants to merge 5 commits into
Open
Conversation
dosubot
Bot
added
the
size:M
baiyutang
force-pushed
the
grafana-dashboard-example
branch
2 times, most recently
from
July 16, 2026 04:47
27c17cb to
8d25da2
Compare
Adds an importable dashboard covering gen_ai_* metrics: total requests and tokens, TTFT and request duration (as quick-glance stats and as percentile trends), token usage by type, original-vs-response model mismatches, request throughput (QPS), token consumption rate by type, overall error rate, and success rate by model. Also adds a README section explaining how to import the dashboard and validate it locally with aigw run. The success rate query uses an `or` fallback so models with a 100% failure rate (no matching series in the failure-count vector) still show up in the table instead of being silently dropped. Suggested in the 2026-07-13 community meeting; unclaimed until now. Signed-off-by: baiyutang <irichardwang@gmail.com>
baiyutang
force-pushed
the
grafana-dashboard-example
branch
from
July 16, 2026 04:50
8d25da2 to
f8249e6
Compare
Member
|
Thanks! For the requests by model and errors by model, can we represent the table differently, or at least remove the timestamp? That timestamp in the table doesn't seem meaningful? |
The instant queries backing the two table panels (requests by model, success rate by model) always surface a Time column with the query timestamp repeated on every row. Add an Organize-fields transformation to drop it and rename the Value column to something readable. Addresses review feedback from @nacx. Signed-off-by: baiyutang <irichardwang@gmail.com>
Member
|
Can you reshare a screenshot with the update? |
missBerg
added
documentation
Author
done |
nacx
enabled auto-merge (squash)
July 22, 2026 00:22
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## main #2374 +/- ##
=======================================
Coverage 85.03% 85.03%
=======================================
Files 159 159
Lines 22634 22634
=======================================
Hits 19247 19247
Misses 2219 2219
Partials 1168 1168 ☔ View full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Description
Adds an importable dashboard covering
gen_ai_*metrics: total requests and tokens, TTFT and request duration (as quick-glance stats and as percentile trends), token usage by type, original-vs-response model mismatches, request throughput (QPS), token consumption rate by type, overall error rate, and success rate by model. Also adds a README section explaining how to import the dashboard and validate it locally withaigw run.The success rate query uses an
orfallback so models with a 100% failure rate (no matching series in the failure-count vector) still show up in the table instead of being silently dropped.Related Issues/PRs (if applicable)
Special notes for reviewers (if applicable)
Screenshot
