[
https://issues.apache.org/jira/browse/KAFKA-9198?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16975770#comment-16975770
]
ASF GitHub Bot commented on KAFKA-9198:
---------------------------------------
hachikuji commented on pull request #7701: KAFKA-9198; Complete purgatory
operations on receiving StopReplica
URL: https://github.com/apache/kafka/pull/7701
Force completion of delayed operations when receiving a StopReplica request.
In the case of a partition reassignment, we cannot rely on receiving a
LeaderAndIsr request in order to complete these operations because the leader
may no longer be a replica. Previously when this happened, the delayed
operations were left to eventually timeout.
### Committer Checklist (excluded from commit message)
- [ ] Verify design and implementation
- [ ] Verify test coverage and CI build status
- [ ] Verify documentation (including upgrade notes)
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
> StopReplica handler should complete pending purgatory operations
> ----------------------------------------------------------------
>
> Key: KAFKA-9198
> URL: https://issues.apache.org/jira/browse/KAFKA-9198
> Project: Kafka
> Issue Type: Bug
> Reporter: Jason Gustafson
> Assignee: Jason Gustafson
> Priority: Major
>
> When a reassignment completes, the current leader may need to be shutdown
> with a StopReplica request. It may still have fetch/produce requests in
> purgatory when this happens. We do not have logic currently to force
> completion of these requests which means they are doomed to eventually
> timeout. This is mostly an issue for produce requests which use the default
> request timeout of 30s.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
