tests: assert histogram/exp_histogram temporality survives in_opentelemetry -> out_opentelemetry round trip

[truongnht]

Summary

The actual fix (fluent/cmetrics#279) is already included via the lib: cmetrics: upgrade to v2.2.0 bump merged into master. This PR is now just the missing test coverage.

tests/integration/scenarios/in_opentelemetry/tests/test_in_opentelemetry_001.py::test_opentelemetry_to_opentelemetry_histogram_and_gauge_metrics already sends a histogram with aggregation_temporality: 1 (DELTA) through a real in_opentelemetry → out_opentelemetry pipeline (test_metrics_002.in.json), but never asserted on the output's aggregationTemporality field. That gap is exactly how the cmetrics bug — histograms/exp_histograms losing their OTLP temporality across Fluent Bit's internal msgpack buffer — went unnoticed. This adds the missing assertion so a regression here is caught going forward.

Background

Original bug: pack_header() in cmetrics' cmt_encode_msgpack.c only wrote the aggregation_type meta field for CMT_COUNTER. Histograms and exp_histograms decoded from OTLP by in_opentelemetry, buffered internally as msgpack, and re-encoded to OTLP by out_opentelemetry lost their temporality in that round trip. Downstream OTLP receivers that validate temporality (e.g. Prometheus with otlp-deltatocumulative enabled) rejected the metric with invalid temporality and type combination. Filed as #278, fixed in fluent/cmetrics#279.

Test plan

• Built Fluent Bit with lib/cmetrics reverted to pre-v2.2.0 (unpatched): confirmed test_opentelemetry_to_opentelemetry_histogram_and_gauge_metrics fails — the response's histogram has no aggregationTemporality field at all (proto JSON omits UNSPECIFIED)
• Built Fluent Bit with lib/cmetrics at v2.2.0 (patched): confirmed the same test passes, with aggregationTemporality: AGGREGATION_TEMPORALITY_DELTA correctly present in the OTLP output
• Ran the full in_opentelemetry integration suite (62 tests) — all pass, no regressions
• Ran the full out_opentelemetry integration suite (17 tests) — all pass, no regressions
• Also reproduced end-to-end against a real cluster: ran a k6 OTLP load test through in_opentelemetry → out_opentelemetry → Prometheus with the unpatched image (k6's histogram metrics rejected by Prometheus) and again with a patched image (same metrics landed correctly, confirmed via direct Prometheus API query)

Summary by CodeRabbit

• Tests
  • Added validation that histogram aggregation temporality remains unchanged during the OpenTelemetry round trip.
  • Confirmed delta temporality is correctly decoded and re-encoded.

[coderabbitai]

📝 Walkthrough

Walkthrough

The OpenTelemetry integration test now verifies that histogram aggregation temporality remains AGGREGATION_TEMPORALITY_DELTA through the Fluent Bit round trip.

Changes

Histogram temporality validation

Layer / File(s)	Summary
Validate histogram temporality tests/integration/scenarios/in_opentelemetry/tests/test_in_opentelemetry_001.py	Adds an assertion covering histogram temporality after decoding, encoding, and the internal msgpack round trip.

Estimated code review effort: 1 (Trivial) | ~2 minutes

Suggested reviewers: cosmo0920

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly describes the added regression test for histogram temporality surviving the OpenTelemetry round trip.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[truongnht]

before the fix

patch fluent-operator with the fix image

[truongnht]

Rebased onto master — the cmetrics fix landed via the lib: cmetrics: upgrade to v2.2.0 bump (3ceed7d), so this PR now only adds the missing test assertion that would have caught the regression. Updated title/description accordingly.