Release Notes - Kafka - Version 1.1.1

Below is a summary of the JIRA issues addressed in the 1.1.1 release of Kafka. For full documentation of the release, a guide to get started, and information about the project, see the Kafka project site.

Note about upgrades: Please carefully review the upgrade documentation for this release thoroughly before upgrading your cluster. The upgrade notes discuss any critical information about incompatibilities and breaking changes, performance changes, and any other changes that might impact your production deployment of Kafka.

The documentation for the most recent release can be found at http://kafka.apache.org/documentation.html.

Improvement

• [KAFKA-6106] - Postpone normal processing of tasks within a thread until restoration of all tasks have completed
• [KAFKA-6760] - responses not logged properly in controller
• [KAFKA-6772] - Broker should load credentials from ZK before requests are allowed
• [KAFKA-6930] - Update KafkaZkClient debug log

Bug

• [KAFKA-6052] - Windows: Consumers not polling when isolation.level=read_committed
• [KAFKA-6054] - ERROR "SubscriptionInfo - unable to decode subscription data: version=2" when upgrading from 0.10.0.0 to 0.10.2.1
• [KAFKA-6566] - SourceTask#stop() not called after exception raised in poll()
• [KAFKA-6661] - Sink connectors that explicitly 'resume' topic partitions can resume a paused task
• [KAFKA-6711] - GlobalStateManagerImpl should not write offsets of in-memory stores in checkpoint file
• [KAFKA-6728] - Kafka Connect Header Null Pointer Exception
• [KAFKA-6739] - Down-conversion fails for records with headers
• [KAFKA-6742] - TopologyTestDriver error when dealing with stores from GlobalKTable
• [KAFKA-6747] - kafka-streams Invalid transition attempted from state READY to state ABORTING_TRANSACTION
• [KAFKA-6748] - Scheduler cannot be cancelled from Punctuator
• [KAFKA-6749] - TopologyTestDriver fails when topoloy under test uses EXACTLY_ONCE
• [KAFKA-6752] - Unclean leader election metric no longer working
• [KAFKA-6765] - Intermittent test failure in CustomQuotaCallbackTest
• [KAFKA-6782] - GlobalKTable GlobalStateStore never finishes restoring when consuming aborted messages
• [KAFKA-6809] - connections-created metric does not behave as expected
• [KAFKA-6825] - DEFAULT_PRODUCTION_EXCEPTION_HANDLER_CLASS_CONFIG is private
• [KAFKA-6853] - ResponseMetadata calculates latency incorrectly (and therefore ZooKeeperRequestLatencyMs is incorrect)
• [KAFKA-6855] - Kafka fails to start with Java 10 due to faulty Java version detection
• [KAFKA-6857] - LeaderEpochFileCache.endOffsetFor() should check for UNDEFINED_EPOCH explicitly
• [KAFKA-6860] - NPE when reinitializeStateStores with eos enabled
• [KAFKA-6870] - Concurrency conflicts in SampledStat
• [KAFKA-6878] - NPE when querying global state store not in READY state
• [KAFKA-6879] - Controller deadlock following session expiration
• [KAFKA-6893] - Processors created after acceptor started which can cause in a brief refusal to accept connections
• [KAFKA-6906] - Kafka Streams does not commit transactions if data is produced via wall-clock punctuation
• [KAFKA-6925] - Memory leak in org.apache.kafka.streams.processor.internals.StreamThread$StreamsMetricsThreadImpl
• [KAFKA-6937] - In-sync replica delayed during fetch if replica throttle is exceeded
• [KAFKA-6975] - AdminClient.deleteRecords() may cause replicas unable to fetch from beginning
• [KAFKA-6982] - java.lang.ArithmeticException: / by zero
• [KAFKA-7012] - Performance issue upgrading to kafka 1.0.1 or 1.1
• [KAFKA-7021] - Source KTable checkpoint is not correct
• [KAFKA-7047] - Connect isolation whitelist does not include SimpleHeaderConverter
• [KAFKA-7058] - ConnectSchema#equals() broken for array-typed default values
• [KAFKA-7082] - Concurrent createTopics calls may throw NodeExistsException
• [KAFKA-7104] - ReplicaFetcher thread may die because of inconsistent log start offset in fetch response
• [KAFKA-7136] - PushHttpMetricsReporter may deadlock when processing metrics changes

Task

• [KAFKA-6854] - Log cleaner fails with transaction markers that are deleted during clean
• [KAFKA-6911] - Incorrect check for keystore/truststore dynamic update
• [KAFKA-6917] - Request handler deadlocks attempting to acquire group metadata lock