dajac commented on code in PR #19611:
URL: https://github.com/apache/kafka/pull/19611#discussion_r2095270860
##########
group-coordinator/src/main/java/org/apache/kafka/coordinator/group/modern/SubscribedTopicDescriberImpl.java:
##########
@@ -79,49 +81,65 @@ public int numPartitions(Uuid topicId) {
*/
@Override
public Set<String> racksForPartition(Uuid topicId, int partition) {
+ TopicImage topic = metadataImage.topics().getTopic(topicId);
+ if (topic != null) {
+ PartitionRegistration partitionRegistration =
topic.partitions().get(partition);
+ if (partitionRegistration != null) {
+ Set<String> racks = new HashSet<>();
+ for (int replica : partitionRegistration.replicas) {
+ // Only add the rack if it is available for the
broker/replica.
+
metadataImage.cluster().broker(replica).rack().ifPresent(racks::add);
Review Comment:
In KIP-881, we don't check whether replicas are online or offline. We
basically assume that they will eventually come back. I would follow the same
approach here.
Side note, this needs to be aligned with the computation of the hash. There,
we include all replicas regardless of whether they are online or not.
--
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
