DanielLeens commented on PR #12305: URL: https://github.com/apache/seatunnel/pull/12305#issuecomment-6289444542
Thanks @SEZ9 for the follow-up, and @goutamadwant for your patience. I can now close the CI question I left open in my review, because the fork run is readable. I earlier guessed that an existing test might assert the old, under-counted byte values. That guess was wrong. The fork run for head `3f4f7ff11c` (Build -> run 39008515550) has exactly one failing job, `engine-v2-it (11, ubuntu-latest)`, and exactly one failing test: ```text BackpressureSlowSinkIT.testCheckpointsKeepCompletingUnderSustainedBackpressure:247 expected at least 3 additional checkpoints to complete during the 90s sustained backpressure window, only observed 2 (samples=[1, 1, 1, 2, 2, 2, 3, 3, 3]) ``` That test is about checkpoint scheduling under a slow sink, and it does not touch `SeaTunnelRow` byte accounting. The same checkpoint-lock-starvation flake is already tracked in #12316 and #12313, and neither is merged to `dev` yet, so syncing `dev` will not clear it. My read is that this failure is unrelated to this PR. That is an inference from the test name, the assertion and the open tracking PRs, not something this PR's diff can influence. One more thing to know: when that job failed, GitHub cancelled the rest of the run. The other jobs (`all-connectors-it-*`, `jdbc-connectors-it-*`, `kafka`, `iceberg` and others) show as cancelled, not passed. So the Build is red, but a clean run has not yet exercised those jobs against this change. Suggested next step for the author: use "Re-run failed jobs" on that fork run (or push a sync with `dev`, which also triggers a fresh run). If `BackpressureSlowSinkIT` is the only red job again, it can be treated as the known flake. If any other job fails, please paste the job and test name here and I will look at it. This does not change the source-side picture from my review: no blocking issues in the code. The compatibility note in the PR description that @SEZ9 asked for is still a nice-to-have. -- 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]
