KAFKA-19631 Admin#describeReplicaLogDirs does not handle error correctly #20461
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
frankvicky
reviewed
Sep 4, 2025
| } | ||
|
|
||
| if (!pendingPartitions.isEmpty() && !directoryFailures.isEmpty()) { | ||
| ArrayList<String> errorAtDir = new ArrayList<>(); |
There was a problem hiding this comment.
nit
Suggested change
| ArrayList<String> errorAtDir = new ArrayList<>(); | |
| List<String> errorAtDir = new ArrayList<>(); |
Comment on lines
3153
to
3155
| for (Map.Entry<String, Throwable> entry : directoryFailures.entrySet()) { | ||
| errorAtDir.add(entry.getValue().getClass().getName() + " at " + entry.getKey()); | ||
| } |
There was a problem hiding this comment.
nit:
Suggested change
| for (Map.Entry<String, Throwable> entry : directoryFailures.entrySet()) { | |
| errorAtDir.add(entry.getValue().getClass().getName() + " at " + entry.getKey()); | |
| } | |
| directoryFailures.forEach((k, v) -> errorAtDir.add(v.getClass().getName() + " at " + k)); |
Comment on lines
2534
to
2535
| Throwable e = assertThrows(Exception.class, () -> values.get(tpr).get()); | ||
| assertInstanceOf(ClusterAuthorizationException.class, e.getCause()); |
There was a problem hiding this comment.
Please use TestUtils#assertFutureThrows instead.
| prepareDescribeLogDirsResponse(Errors.CLUSTER_AUTHORIZATION_FAILED, broker1log0), | ||
| env.cluster().nodeById(tpr.brokerId())); | ||
|
|
||
| DescribeReplicaLogDirsResult result = env.adminClient().describeReplicaLogDirs(singletonList(tpr)); |
| DescribeLogDirsResponseData.DescribeLogDirsResult failedResult = new DescribeLogDirsResponseData.DescribeLogDirsResult() | ||
| .setErrorCode(Errors.LOG_DIR_NOT_FOUND.code()) | ||
| .setLogDir(broker1log1); | ||
| DescribeLogDirsResponse response = new DescribeLogDirsResponse(new DescribeLogDirsResponseData().setResults(asList(successfulResult, failedResult))); |
| DescribeLogDirsResponse response = new DescribeLogDirsResponse(new DescribeLogDirsResponseData().setResults(asList(successfulResult, failedResult))); | ||
| env.kafkaClient().prepareResponseFrom(response, env.cluster().nodeById(successfulTpr.brokerId())); | ||
|
|
||
| DescribeReplicaLogDirsResult result = env.adminClient().describeReplicaLogDirs(asList(successfulTpr, failedTpr)); |
| Map<TopicPartitionReplica, KafkaFuture<DescribeReplicaLogDirsResult.ReplicaLogDirInfo>> values = result.values(); | ||
|
|
||
| assertNotNull(values.get(successfulTpr).get()); | ||
| assertThrows(Exception.class, () -> values.get(failedTpr).get()); |
There was a problem hiding this comment.
Do you think we should assert the class of this exception?
There was a problem hiding this comment.
Yes, an assertion on the exception type and could better ensure that the future ends with the expected reason. They are included in the new commit.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
Admin#desceibeReplicaLogDirs fails all futures when any error happens, regardless of the exception type or the failed directory. Especially, required by the issue, CLUSTER_AUTHORIZATION_FAILED should be handled independently.
Changes
Tests
Two tests are added: one for checking whether user can catch the CLUSTER_AUTHORIZATION_FAILED by getCause; one for checking whether failure in unrelated directory will affect other requests.
CLUSTER_AUTHORIZATION_FAILED
It is handled by an additional if statement now.
Single Directory-level Error
It is hard to map failed directories back to requested topic partition replica. If a directory or a broker fails, the response won't contain information of any replicas. The new attempt fails replicas that are not mentioned in healthy responses, if there is some error codes. It won't break the behavior of giving an empty result to unreplied replicas if no error code is in response.