jolshan commented on a change in pull request #10952: URL: https://github.com/apache/kafka/pull/10952#discussion_r669899334
########## File path: clients/src/main/java/org/apache/kafka/clients/MetadataCache.java ########## @@ -130,13 +150,30 @@ MetadataCache mergeWith(String newClusterId, Set<String> addInvalidTopics, Set<String> addInternalTopics, Node newController, + Map<String, Uuid> topicIds, BiPredicate<String, Boolean> retainTopic) { Predicate<String> shouldRetainTopic = topic -> retainTopic.test(topic, internalTopics.contains(topic)); Map<TopicPartition, PartitionMetadata> newMetadataByPartition = new HashMap<>(addPartitions.size()); + Map<String, Uuid> newTopicIds = new HashMap<>(topicIds.size()); + + // We want the most recent topic ID. We add the old one here for retained topics and then update with newest information in the MetadataResponse + // we add if a new topic ID is added or remove if the request did not support topic IDs for this topic. + for (Map.Entry<String, Uuid> entry : this.topicIds.entrySet()) { + if (shouldRetainTopic.test(entry.getKey())) { + newTopicIds.put(entry.getKey(), entry.getValue()); + } + } + for (PartitionMetadata partition : addPartitions) { newMetadataByPartition.put(partition.topicPartition, partition); + Uuid id = topicIds.get(partition.topic()); + if (id != null) + newTopicIds.put(partition.topic(), id); + else + // Remove if the latest metadata does not have a topic ID Review comment: Yeah. That was my reasoning. I thought the upgrade/downgrade case would be rare and the guarantees harder to reason about there. -- 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