SEZ9 commented on issue #11437: URL: https://github.com/apache/seatunnel/issues/11437#issuecomment-5155772027
Thanks for the detailed report. The 'Steps to reproduce' section appears empty/truncated — could you paste the exact job config, cluster hazelcast/seatunnel.yaml (slot and schedule settings), and the sequence of commands so we can reproduce on dev commit d6df7f2f? Please attach Master and Worker logs around the failover window (09:38:34Z ± a few minutes), especially ResourceManager/SlotService entries on the new active Master and both Workers, to confirm whether the new Master ever re-synced the Workers' slot assignment state after re-election. The symptom suggests two related defects: (1) after failover the new Master cannot reuse or reclaim slots still marked assigned to the restoring job, so allocation blocks pre-deployment; (2) job cancellation does not release Worker-held slots. Can you confirm whether the Workers' SlotService reported any releaseSlot/heartbeat activity for the old job ID after cancel? This will help decide if it's a ResourceManager state-restore gap versus a Worker-side release path bug. <!-- streview-comment:38 --> -- 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]
