lhotari commented on PR #23874: URL: https://github.com/apache/pulsar/pull/23874#issuecomment-2607267800
> I think we'd better solve the issue this way, which has less lock range, better performance, and easier to read @poorbarcode This looks like a useful optimization, but I don't think that it addresses the reported issue #23870. In that case, the problem occurs at a different stage. In the issue, it's an ordinary read that is happening and when the hash gets unblocked, this is the line where the hash is no longer blocked: https://github.com/apache/pulsar/blob/e5bd77419e91d1602731cd0c1d02a738e1b7ebc7/pulsar-broker/src/main/java/org/apache/pulsar/broker/service/persistent/PersistentStickyKeyDispatcherMultipleConsumers.java#L510 The last acknowledgement will remove the hash from unblocking before it has been added to the replay queue. The alternative solution for this PR would be to postpone the removal after the message has been added to the replay queue. -- 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]
