[songgaoye]

What problem does this PR solve?

Issue Number: resolve

Problem Summary:

RecorderTest.latency_recorder_qps_accuracy is flaky because it calculates the expected QPS from one window sample and then reads qps() separately. A sampler tick can occur between those reads, causing them to refer to different windows and resulting in intermittent zero or inaccurate QPS assertions.

What is changed and the side effects?

Changed:

• Retry the QPS assertion when the latency-window sample changes between reads.
• Compare qps() with the expected value only when the samples immediately before and after the QPS read have the same timestamp.

Side effects:

• Performance effects: Test-only change. Negligible additional polling in the rare case of a sampler tick during validation.
• Breaking backward compatibility: None.

---

Check List:

• Please make sure your changes are compilable.
• When providing us with a new feature, it is best to add related tests.
• Please follow Contributor Covenant Code of Conduct.

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has insufficient remaining quota.

[Copilot]

🟡 Changes recommended

The current test can validate an unintended 2-second window and can compare QPS values from different samples.

1 open issue

• Use a 3-second recorder for 3-second QPS cases · New

_{Review effort: Lite}

[Copilot]

🟡 Changes recommended

The revised test can still validate incorrect windows and remain flaky under the sampler race.

1 open issue

• Test accepts incomplete sampling windows · New

1 resolved since last review

• Use a 3-second recorder for 3-second QPS cases

_{Review effort: Lite}

[wwbmmm]

[replied by brpc-oncall robot] The rework addresses the previous round (dedicated 2s/3s recorders plus a full-window guard), and Copilot's claim about has_full_latency_window() does not hold; only the one-tick observation interval remains a residual scheduling risk.

[Copilot]

🔵 Needs a closer look

Address the two moderate test-synchronization issues before approval.

No open issues 🎉

1 resolved since last review

• Test accepts incomplete sampling windows

_{Review effort: Lite}