kfaraz commented on PR #19091: URL: https://github.com/apache/druid/pull/19091#issuecomment-4009696005
> In the example above, how are we avoiding the case where we don't actually want to allow B to continue b/c it's using invalid offsets? For example, what if A fails to publish? @jtuglu1 , that case is already handled in the existing code. After A publishes to fail, the code flow moves into the `else` block and we check if the checkpoints of B are consistent with the persisted offsets or not. Since they wouldn't be consistent, B will be killed off and a new task group C will be created. As long as A is pending publish though, we have to assume that it is going to succeed and continue with B as usual. -- 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]
