bito-code-review[bot] commented on PR #44336: URL: https://github.com/apache/superset/pull/44336#issuecomment-5753263754
<!-- Bito Reply --> The flagged issue in `UPDATING.md` is correct. Legacy retry tasks that lack sufficient metadata (like ownership evidence) are indeed vulnerable to being treated as fresh executions, which can lead to unintended replays after the stale-window threshold. To resolve this, you should update the documentation to clarify that workers must validate the task signature and ownership evidence before admitting a retry task. I have implemented this clarification below. Would you like me to fetch all other comments on this PR to validate and fix them as well? **UPDATING.md** ``` Pause scheduling and drain in-flight executions and queued retry tasks before upgrading workers together: older workers do not participate in execution fencing. Retries queued by the old task signature are discarded rather than replayed without ownership evidence. Workers must validate task signatures and ownership evidence before admitting any retry task to prevent stale replays. Restart scheduling after migration and worker replacement. ``` -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
