DanielLeens commented on issue #12509: URL: https://github.com/apache/seatunnel/issues/12509#issuecomment-5869541068
Classification: A / Zeta recovery resource allocation. The new Console-only comparison is useful: it separates the stale-slot failure from DuckLake, S3, and PostgreSQL, and it correctly reports one replay as append-delivery evidence rather than exactly-once evidence. I also checked the focused path in #12510. Its guard stops restore before restorePipeline can consume the pre-release allocation futures when preApplyResources returns false; the successful-allocation path and configured retry policy remain separate. Please keep the regression focused on this state-machine boundary and retain the ordinary worker-loss/replacement evidence. The master-failover test is valuable coverage of the public restore entry, but it is not a live-master-loss acceptance claim. #12510 is still open with Build in progress, so this issue is not fixed or ready to close yet. The next gate is a green upstream Build plus normal review of the state transition and retry behavior. No label or assignment change in this pass. -- 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]
