kevin-wu24 commented on code in PR #19668:
URL: https://github.com/apache/kafka/pull/19668#discussion_r2114517892
##########
raft/src/test/java/org/apache/kafka/raft/KafkaRaftClientReconfigTest.java:
##########
@@ -2250,13 +2250,16 @@ void
testFollowerSendsUpdateVoterWithKraftVersion0(Errors updateVoterError) thro
new LeaderAndEpoch(OptionalInt.of(voter1.id()), epoch)
)
);
+ // polling sends a fetch because no fetches are in flight, only the
update voter
context.client.poll();
+ RaftRequest.Outbound fetchRequest = context.assertSentFetchRequest();
+ context.assertFetchRequestData(fetchRequest, epoch, 0L, 0);
Review Comment:
We need to perform this poll here in order to make sure the local replica
does not expire its fetch timeout and transition to prospective.
This is because the poll will first send a fetch since there is not one in
flight and its fetch timeout is not expired, and then handle the incoming
`UpdateVoterResponse`, which will reset the fetch timeout.
Calling `context.pollUntilRequest()` below is too late because the fetch
timeout will have expired by that point. This is because earlier, we slept for
`context.fetchTimeoutMs - 1` BEFORE completing an `updateVoter` RPC (i.e. there
is 1 ms left on the fetch timeout at this point).
--
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
