lqjacklee created KAFKA-10208: --------------------------------- Summary: org.apache.kafka.clients.consumer.internals.ConsumerCoordinator.OffsetFetchResponseHandler return null when Broker unexpectedly doesn't support requireStable flag on version while not any information Key: KAFKA-10208 URL: https://issues.apache.org/jira/browse/KAFKA-10208 Project: Kafka Issue Type: Bug Components: clients Affects Versions: 2.7.0 Reporter: lqjacklee
When the 2.7.0 client try to request the broker whose version is 2.3.0, the OffsetAndMetadata will be null and miss the Key information. I have create the test case as below : @Test public void testCreateTopicAndCheckTheOffsite() throws ExecutionException, InterruptedException { String topicName = UUID.randomUUID().toString(); String groupId = "DEMO_" + topicName; final Properties props = new Properties(); props.put(StreamsConfig.APPLICATION_ID_CONFIG, groupId); props.put(StreamsConfig.BOOTSTRAP_SERVERS_CONFIG, "localhost:9092"); props.put(StreamsConfig.CACHE_MAX_BYTES_BUFFERING_CONFIG, 0); String serializer = StringSerializer.class.getName(); String deserializer = StringDeserializer.class.getName(); props.put("auto.offset.reset", "latest"); props.put("session.timeout.ms", "30000"); props.put("key.deserializer", deserializer); props.put("value.deserializer", deserializer); props.put("key.serializer", serializer); props.put("value.serializer", serializer); props.put(ConsumerConfig.REQUEST_TIMEOUT_MS_CONFIG, 30000); AdminClient adminClient = AdminClient.create(props); boolean topicExist = false; try { NewTopic newTopic = new NewTopic(topicName, 1, (short) 1); CreateTopicsOptions createTopicsOptions = new CreateTopicsOptions(); createTopicsOptions.timeoutMs(3000000); final CreateTopicsResult createTopicsResult = adminClient.createTopics(Collections.singleton(newTopic), createTopicsOptions); createTopicsResult.values().get(topicName).get(); }catch (TopicExistsException e) { topicExist = true; } try { List<TopicPartition> topicPartitions = new ArrayList<>(); KafkaConsumer<String,String> kafkaConsumer = new KafkaConsumer<>(props); Field kafkaClientField = kafkaConsumer.getClass().getDeclaredField("client"); kafkaClientField.setAccessible(true); ConsumerNetworkClient client = (ConsumerNetworkClient) kafkaClientField.get(kafkaConsumer); FindCoordinatorRequest.Builder findCoordinatorRequest = new FindCoordinatorRequest.Builder( new FindCoordinatorRequestData() .setKeyType(FindCoordinatorRequest.CoordinatorType.GROUP.id()) .setKey(groupId)); Node node = client.leastLoadedNode(); Node coordinator; RequestFuture<Node> requestCoordinatorFuture = client.send(node, findCoordinatorRequest) .compose(new RequestFutureAdapter<ClientResponse, Node>() { @Override public void onFailure(RuntimeException e, RequestFuture<Node> future) { super.onFailure(e, future); } @Override public void onSuccess(ClientResponse value, RequestFuture<Node> future) { Node coordinator; FindCoordinatorResponse findCoordinatorResponse = (FindCoordinatorResponse) value.responseBody(); Errors error = findCoordinatorResponse.error(); if (error == Errors.NONE) { // use MAX_VALUE - node.id as the coordinator id to allow separate connections // for the coordinator in the underlying network client layer int coordinatorConnectionId = Integer.MAX_VALUE - findCoordinatorResponse.data().nodeId(); coordinator = new Node( coordinatorConnectionId, findCoordinatorResponse.data().host(), findCoordinatorResponse.data().port()); client.tryConnect(coordinator); future.complete(coordinator); } else if (error == Errors.GROUP_AUTHORIZATION_FAILED) { Assert.fail(error.message()); } else { future.raise(error); } } }); client.poll(requestCoordinatorFuture); if (requestCoordinatorFuture.succeeded()) { coordinator = requestCoordinatorFuture.value(); } else { throw requestCoordinatorFuture.exception(); } OffsetFetchRequest.Builder requestBuilder = new OffsetFetchRequest.Builder(groupId, true, topicPartitions, true); RequestFuture<Map<TopicPartition, OffsetAndMetadata>> topicPartitionMetadataRequestFuture = client.send(coordinator, requestBuilder) .compose(new RequestFutureAdapter<ClientResponse, Map<TopicPartition, OffsetAndMetadata>>() { @Override public void onSuccess(ClientResponse value, RequestFuture<Map<TopicPartition, OffsetAndMetadata>> future) { OffsetFetchResponse response = (OffsetFetchResponse) value.responseBody(); if (response.hasError()) { Errors error = response.error(); if (error == Errors.COORDINATOR_LOAD_IN_PROGRESS) { // just retry future.raise(error); } else if (error == Errors.NOT_COORDINATOR) { // re-discover the coordinator and retry future.raise(error); } else if (error == Errors.GROUP_AUTHORIZATION_FAILED) { Assert.fail(Errors.GROUP_AUTHORIZATION_FAILED + ""); } else { future.raise(new KafkaException("Unexpected error in fetch offset response: " + error.message())); } return; } Set<String> unauthorizedTopics = null; Map<TopicPartition, OffsetAndMetadata> offsets = new HashMap<>(response.responseData().size()); Set<TopicPartition> unstableTxnOffsetTopicPartitions = new HashSet<>(); for (Map.Entry<TopicPartition, OffsetFetchResponse.PartitionData> entry : response.responseData().entrySet()) { TopicPartition tp = entry.getKey(); OffsetFetchResponse.PartitionData partitionData = entry.getValue(); if (partitionData.hasError()) { Errors error = partitionData.error; if (error == Errors.UNKNOWN_TOPIC_OR_PARTITION) { future.raise(new KafkaException("Topic or Partition " + tp + " does not exist")); return; } else if (error == Errors.TOPIC_AUTHORIZATION_FAILED) { if (unauthorizedTopics == null) { unauthorizedTopics = new HashSet<>(); } unauthorizedTopics.add(tp.topic()); } else if (error == Errors.UNSTABLE_OFFSET_COMMIT) { unstableTxnOffsetTopicPartitions.add(tp); } else { future.raise(new KafkaException("Unexpected error in fetch offset response for partition " + tp + ": " + error.message())); return; } } else if (partitionData.offset >= 0) { // record the position with the offset (-1 indicates no committed offset to fetch); // if there's no committed offset, record as null offsets.put(tp, new OffsetAndMetadata(partitionData.offset, partitionData.leaderEpoch, partitionData.metadata)); } else { try { HashMap<TopicPartition, OffsetSpec> offsetMap = new HashMap<>(); offsetMap.put(tp, OffsetSpec.earliest()); ListOffsetsResult listOffsetsResult = adminClient.listOffsets(offsetMap); Map<TopicPartition, ListOffsetsResult.ListOffsetsResultInfo> topicPartitionListOffsetsResultInfoMap = listOffsetsResult.all().get(); ListOffsetsResult.ListOffsetsResultInfo offsetsResultInfo = topicPartitionListOffsetsResultInfoMap.get(tp); offsets.put(tp, new OffsetAndMetadata(offsetsResultInfo.offset(), offsetsResultInfo.leaderEpoch(), "")); } catch (Exception e) { Assert.fail(e.getMessage()); } } Assert.fail("not found the topic and partition"); } if (unauthorizedTopics != null) { future.raise(new TopicAuthorizationException(unauthorizedTopics)); } else if (!unstableTxnOffsetTopicPartitions.isEmpty()) { // just retry future.raise(new UnstableOffsetCommitException("There are unstable offsets for the requested topic partitions")); } else { future.complete(offsets); } } @Override public void onFailure(RuntimeException e, RequestFuture<Map<TopicPartition, OffsetAndMetadata>> future) { super.onFailure(e, future); } }); client.poll(topicPartitionMetadataRequestFuture); if(topicPartitionMetadataRequestFuture.succeeded()) { Map<TopicPartition, OffsetAndMetadata> value = topicPartitionMetadataRequestFuture.value(); Assert.assertNotNull(value); }else { Assert.fail(topicPartitionMetadataRequestFuture.exception().getMessage()); } }catch (Exception e) { Assert.fail(e.getMessage()); }finally { if(topicExist) { List<String> topicToDeleted = new ArrayList<>(); DeleteTopicsResult deleteTopicsResult = adminClient.deleteTopics(topicToDeleted); deleteTopicsResult.all().get(); } } } -- This message was sent by Atlassian Jira (v8.3.4#803005)