DanielLeens commented on PR #12134: URL: https://github.com/apache/seatunnel/pull/12134#issuecomment-5717371529
Pushed `872701e`: this PR's own `CheckpointCoordinatorFailoverIT#testStreamJobFailsAfterCheckpointTriggerDispatchFailure` hits the identical harmless race independently found and fixed in #12288 — `CheckpointManager#sendOperationToMemberNode` backs both the barrier-dispatch and the post-completion notify call, so the injected fault can surface via either path depending on scheduling. Widened the assertion to accept both `CHECKPOINT_INSIDE_ERROR` and `CHECKPOINT_NOTIFY_COMPLETE_FAILED`, same as #12288's interim fix. Note: #12288 later replaced that with a fully deterministic fix backed by a new `CheckpointCoordinator#getLatestCompletedCheckpointId()` src/main method — porting that here would mean duplicating an engine-side change into a PR that's otherwise about the cancel/savepoint-failover race, so I left it at the test-only widening pending a decision on whether to do that instead. -- 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]
