DanielLeens commented on PR #12173: URL: https://github.com/apache/seatunnel/pull/12173#issuecomment-5696057503
@Rangsh @SEZ9 — re-confirmed at `7948e2640` (no new push since my last check). Rather than taking the status line at face value, I pulled the PR head directly (`git fetch origin refs/pull/12173/head` → `7948e26402317fb8cf4bba22ad4b1f924e3436a8`, matches) and checked each point against source: - **F4 (docs)**: this head's file list (`IMapJobStorageBenchmark.java`, `IMapJobGrowthBenchmarkWorkload.java`, `JobHistoryService.java`, `JobHistoryServiceFinishedMetricsTest.java`) no longer touches `docs/en|zh/engines/zeta/benchmark.md` at all — the notes added in `909f878` and removed in `9a6077c58` net out to zero diff vs `dev`, and current `dev`'s `benchmark.md` has no `runningJobGrowth`/`completedJobHistoryGrowth` mentions either. Removal confirmed. - **F1/F3**: `IMapJobGrowthBenchmarkWorkload.shouldSampleGrowthDurability()` gates on `growthIterationIndex == 0 && initialStoredJobCount == 0`, and `@TearDown(Level.Trial)` calls `verifyGrowthBatchDurability()` unconditionally. So `count=0` samples first-iteration + trial-teardown, `count=1000` samples trial-teardown only — matches what was described. The resident-only gap is documented in the `verifyGrowthPhase()` Javadoc. - **F2/F5**: `JobHistoryService.storeFinishedPipelineMetrics` wraps `get → merge → put` in `finishedJobMetricsImap.lock(jobId)` / `unlock(jobId)` (`try`/`finally`); Javadoc is contract-only, no "in memory" or historical `computeIfAbsent` wording left. - **F6**: `JobHistoryServiceFinishedMetricsTest` has `storeFinishedPipelineMetricsRejectsNullMetrics`, and the new-job/merge/distinct-tag tests all assert the TTL-preserving `put(..., 1440, MINUTES)` plus explicit `lock`/`unlock` verification. All four check out against the actual diff, not just the summary — nothing reopened on my side, F1–F6 stay closed, and this remains a status-confirmation round rather than a new review pass since there's still no commit past `7948e2640`. The one live gap is unchanged: `reviewDecision` is still `CHANGES_REQUESTED` and `mergeStateStatus` is still `BLOCKED`, both pinned to @nzw921rx's 2026-09-10 review that predates the doc fix in `9a6077c58`. That remains the sole item standing between this PR and merge from where I sit. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
