DanielLeens commented on PR #11841: URL: https://github.com/apache/seatunnel/pull/11841#issuecomment-6396814889
Thanks, @SEZ9. I traced those two questions at `881e275fc22830552d83244ad7322114352c7741` with `git show` and a call-site search: 1. `JobMaster.init` at lines 275-293 explicitly documents and enforces skipping cluster-side save-mode handling for restore jobs. Its normal call therefore uses the non-restore mode. The overload at lines 782-799 also preserves the public boolean helper and delegates to `RestoreMode`; the restore branch is not exercised by that guarded init call. I would keep the existing restore guard: moving it would change restore-time save-mode execution and needs a separate behavioral decision/test. The author can clarify the public-helper intent rather than implying this init path exercises it. 2. The broader flag is passed at `JobMaster.initCheckPointManager`, lines 338-345: `jobImmutableInformation.isRestoreJob() || restart` goes into `CheckpointManager`. Its constructor path then passes that `isRestoreJob` value to `new CheckpointCoordinator(...)` at `CheckpointManager.java:198-209`. That is the intermediate call chain behind the rename; it is not limited to a direct constructor call in `JobMaster`. These are source-level answers to the existing discussion, not a new full review or an executed restore test. Please let me know if a different caller is the concern. -- 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]
