philipnee commented on code in PR #14937:
URL: https://github.com/apache/kafka/pull/14937#discussion_r1418122441
##########
clients/src/main/java/org/apache/kafka/clients/consumer/internals/AsyncKafkaConsumer.java:
##########
@@ -1050,6 +1062,81 @@ private void close(Duration timeout, boolean
swallowException) {
}
}
+ /**
+ * Prior to closing the network thread, we need to make sure the following
operations happen in the right sequence:
+ * 1. autocommit offsets
+ * 2. revoke all partitions
+ */
+ private void prepareShutdown(final Timer timer, final
AtomicReference<Throwable> firstException) {
+ if (!groupMetadata.isPresent())
+ return;
+ maybeAutoCommitSync(timer, firstException);
+ timer.update();
+ waitOnEventCompletion(new
ConsumerCloseApplicationEvent(ConsumerCloseApplicationEvent.Task.COMMIT,
timer.remainingMs()), timer, firstException);
+ maybeInvokeCommitCallbacks();
+ maybeRevokePartitions(timer, firstException);
+ waitOnEventCompletion(new
ConsumerCloseApplicationEvent(ConsumerCloseApplicationEvent.Task.LEAVE_GROUP,
timer.remainingMs()), timer, firstException);
+ }
+
+ private void waitOnEventCompletion(final ConsumerCloseApplicationEvent
event,
+ final Timer timer,
+ final AtomicReference<Throwable>
firstException) {
+ try {
+ applicationEventHandler.addAndGet(event, timer);
+ } catch (TimeoutException e) {
Review Comment:
We don't really throw timeout exceptions during closing because if user
tries to close with 0 duration then all ops will be timedout. The current
implementation just polls, but since we cannot directly polls the client, we
need to either wait till the future is completed or times out and keep going.
--
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
