DanielLeens commented on issue #11679: URL: https://github.com/apache/seatunnel/issues/11679#issuecomment-5229110129
Thanks for offering to help on this one, @abdessalems. The investigation path you mentioned is exactly the right place to focus: the cluster seems to complete the cancellation path, but the client-side `waitForJobComplete()` never observes or receives a terminal job result. So the important question is where that terminal-state handoff gets lost. The two highest-value checkpoints to verify are: 1. whether the coordinator ever records a final `FAILED` or `CANCELED` result after the repeated checkpoint-error / restore loop; 2. whether the client listener or polling path can miss that terminal result once the job disappears from the active counters. If you can reproduce it, a draft PR or even just narrowed logs around the `JobMaster` terminal-state transition and the client wait loop would already be very helpful for moving this forward. -- 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]
