DanielLeens commented on PR #12391: URL: https://github.com/apache/seatunnel/pull/12391#issuecomment-5759083178
Thanks for the clear CI breakdown, @SeaSand1024, and for keeping this PR scoped. I would keep it exactly as it is: please do not fold the #12353 / #12377 cancel-race fix into #12391. That fix touches `SubPlan` terminal-state selection, which is a different contract from the close-all-flow-lifecycles change here, and mixing the two would make both harder to review and to revert. On my side, the source-level points from my last review on `0fc63dc5` are closed, so I do not see a new code-side blocker in this PR. The remaining gate is CI on the fork run. What I can see from the fork run for this head (run attempt 3): the failed jobs are `engine-v2-it` and `all-connectors-it-2` on both JDK 8 and JDK 11, and the unit-test matrix is not among them. That matches your description. I have not independently re-read each individual assertion line in those job logs, so please treat the attribution below as a reading of the job names plus your notes rather than a log-level proof: - `SplitClusterFaultToleranceIT.testStreamJobCancelResolvesWhenWorkerCrashesBeforeCancelAck` (expected `CANCELED`, got `FAILED`) is the known cancel-versus-worker-loss race. It is not in the code this PR changes. Fixes for it are open as #12311 (resolves a CANCELING vertex to CANCELED when its worker is lost) and #12377 (preserve cancellation after worker failure); neither is in `dev` yet, so syncing `dev` will not clear it. - `OpengaussCDCIT` lives in `all-connectors-it-2` and does not touch the engine lifecycle code in this PR. Suggested next step: use "Re-run failed jobs" only, and do not add unrelated changes to this branch. If the same two jobs stay red after a re-run, please paste the new failing job links here and I will take another look at that exact head. A committer can then decide whether to merge on the strength of the green unit-test matrix plus the known-flake attribution, or wait for #12311 / #12377 to land first. -- 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]
