siddhantsangwan commented on PR #5632: URL: https://github.com/apache/ozone/pull/5632#issuecomment-1831421358
> But the quasi-closed which can be forced closed is an interesting problem - should we return true here (we do currently)? What if, the replicas are failing to close somehow, and it is under-replicated by 1. It will never get closed and it will never check for under-replication either. Perhaps we should try to close it, and return false to let the other handlers run too. That does make sense to me, since we allow quasi closed replicas to get replicated. We're also already following a similar pattern of returning false even after handling in `MismatchedReplicasHandler` so that replication can be checked, since handling under replication is higher priority. I wonder why this wasn't being done in the legacy RM for quasi closed containers? It tries to close the container and returns from the method. The new RM adopted this behaviour from there. -- 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]
