hachikuji commented on a change in pull request #11231:
URL: https://github.com/apache/kafka/pull/11231#discussion_r694414167
##########
File path:
clients/src/main/java/org/apache/kafka/clients/consumer/internals/AbstractCoordinator.java
##########
@@ -476,7 +476,6 @@ boolean joinGroupIfNeeded(final Timer timer) {
else if (!future.isRetriable())
throw exception;
- resetStateAndRejoin(String.format("rebalance failed with
retriable error %s", exception));
Review comment:
@guozhangwang @ableegoldman After looking into this issue, my conclusion
is that no additional changes are needed beyond the one line removal here. If
the subscription/metadata changes while the rebalance is in progress, then we
will detect it in `onJoinComplete` and we will call `requestRejoin`. But after
`onJoinComplete` returns, we already reset `needsJoinPrepare`, so we should not
miss the call to `onJoinPrepare` after we send the next `JoinGroup`.
On the other hand, if the rebalance fails because of an error such as
UNKNOWN_MEMBER_ID, then `JoinResponseHandler` already invokes
`resetGenerationOnResponseError`. After the fix from KAFKA-12983, this ensures
that `needsJoinPrepare` gets reset.
Are there any other cases to consider?
--
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: jira-unsubscr...@kafka.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
