plugins(stats_over_http): Add Prometheus v2 output with metric labels #12792
+487
−28
Conversation
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
Contributor
There was a problem hiding this comment.
Pull request overview
This pull request adds a new Prometheus v2 output format for the stats_over_http plugin that transforms flat ATS metric names into labeled metrics for better aggregation in Prometheus/Grafana. The new format can be accessed via /_stats/prometheus_v2 endpoint or by passing Accept: text/plain; version=2.0.0 header.
Key changes:
- Implements metric name parsing logic to extract labels (status codes, HTTP methods, cache results, volume/thread indices, time buckets) from flat metric names
- Adds deduplication of HELP/TYPE metadata lines for metrics sharing the same base name
- Includes compile-time unit tests and comprehensive AuTests for validation
Reviewed changes
Copilot reviewed 5 out of 5 changed files in this pull request and generated 4 comments.
Show a summary per file
| File | Description |
|---|---|
| plugins/stats_over_http/stats_over_http.cc | Core implementation: adds parse_metric_v2 function, prometheus_v2_out_stat callback, new output format enum, HTTP headers, routing logic, and compile-time tests |
| tests/gold_tests/pluginTest/stats_over_http/stats_over_http.test.py | Adds test cases for Prometheus v2 format via both endpoint path and Accept header, with assertions validating label extraction |
| tests/gold_tests/pluginTest/stats_over_http/gold/stats_over_http_prometheus_v2_stderr.gold | Gold file for expected stderr output when accessing /_stats/prometheus_v2 endpoint |
| tests/gold_tests/pluginTest/stats_over_http/gold/stats_over_http_prometheus_v2_accept_stderr.gold | Gold file for expected stderr output when using Accept header for Prometheus v2 |
| doc/admin-guide/plugins/stats_over_http.en.rst | Documentation updates explaining the new v2 format, endpoints, and content-type headers |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
The current Prometheus output in stats_over_http dumps flattened metric names (e.g., proxy.process.http.200_responses), which makes it difficult to aggregate data in Grafana or Prometheus. This adds a new "v2" output format that parses these names into proper labels like status="200" or method="GET". The new format can be accessed via /_stats/prometheus_v2 or by passing an Accept header with version=2.0.0. To keep this efficient, the parsing uses swoc::TextView for a zero- allocation approach. It recognizes common ATS patterns such as status codes, HTTP methods, cache results, and thread/volume indices. I've also added deduplication for the HELP and TYPE metadata lines, so that metrics sharing the same base name (but different labels) are grouped correctly in the output. Includes compile-time unit tests (static_assert) to verify the parsing logic and updated AuTests for the new endpoint.
bneradt
requested changes
Jan 8, 2026
Comment on lines
+109
to
+112
| p.Streams.stdout += Testers.ContainsExpression( | ||
| 'proxy_process_http_requests{method="completed"}', | ||
| "Verify that method labels are extracted correctly.", | ||
| ) |
Contributor
There was a problem hiding this comment.
I'm glad you added this test.
method="completed"
completed isn't a method, POST, GET, PUT, etc., are methods. It seems like having completed would be misleading for users viewing this data.
Contributor
|
In case it's helpful, here's the fedora build failure: |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 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.
The current Prometheus output in stats_over_http dumps flattened metric names (e.g.,
proxy.process.http.200_responses), which makes it difficult to aggregate data in Grafana or Prometheus. This adds a new "v2" output format that parses these names into labels likestatus="200"ormethod="GET".The new format can be accessed via /_stats/prometheus_v2 or by passing an Accept header with version=2.0.0.
To keep this efficient, the parsing uses
swoc::TextViewfor a zero- allocation approach. It recognizes common ATS patterns such as status codes, HTTP methods, cache results, and thread/volume indices. I've also added deduplication for the HELP and TYPE metadata lines, so that metrics sharing the same base name (but different labels) are grouped correctly in the output.Includes compile-time unit tests (static_assert) to verify the parsing logic and updated AuTests for the new endpoint.
Related issue: #12778