SEZ9 commented on issue #12566: URL: https://github.com/apache/seatunnel/issues/12566#issuecomment-5924295771
Thanks for the detailed analysis. To help confirm the async-load behavior you describe, could you share your MapStore configuration and master startup logs from the full restart showing `restoreAllRunningJobFromMasterNodeSwitch()` running while `runningJobInfoIMap` is empty? On point 2, could you confirm whether the job had checkpoint storage enabled and whether the checkpoint state was also persisted across the restart? Knowing whether only `runningJobInfoIMap` survived vs. all persisted state will clarify whether skipping checkpoint restore when `restoreSourceJobId` is null is intended or a gap. If full-cluster-restart recovery is confirmed as in scope, it may be worth splitting this into two separate bug issues (1: EAGER MapStore async load yields an empty `entrySet()` on restore; 2: non-savepoint jobs don't resume from the latest checkpoint on cluster restart) so each can be tracked and fixed independently. <!-- streview-comment:1448 --> -- 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]
