Apache Kafkaに関する非互換を以下に示します。
内容 | 詳細 | JIRA | Summary |
|---|---|---|---|
外部仕様の変更 | コマンド仕様の変更（注1） | KAFKA-8421 | Allow consumer.poll() to return data in the middle of rebalance |
KAFKA-10004 | ConfigCommand fails to find default broker configs without ZK | ||
KAFKA-10018 | Change command line tools from /bin/sh to /bin/bash | ||
KAFKA-10056 | Consumer metadata may use outdated groupSubscription that doesn't contain newly subscribed topics | ||
KAFKA-10123 | Regression resetting offsets in consumer when fetching from old broker | ||
KAFKA-7052 | ExtractField SMT throws NPE - needs clearer error message | ||
KAFKA-8847 | Deprecate and remove usage of supporting classes in kafka.security.auth | ||
KAFKA-8933 | An unhandled SSL handshake exception in polling event - needed a retry logic | ||
KAFKA-9233 | Kafka consumer throws undocumented IllegalStateException | ||
KAFKA-9297 | CreateTopic API do not work with older version of the request/response | ||
KAFKA-9329 | KafkaController::replicasAreValid should return error | ||
KAFKA-9386 | Flaky test AclAuthorizerTest.testHighConcurrencyDeletionOfResourceAcls | ||
KAFKA-9558 | getListOffsetsCalls doesn't update node in case of leader change | ||
KAFKA-9807 | Race condition updating high watermark allows reads above LSO | ||
KAFKA-9815 | Consumer may never re-join if inconsistent metadata is received once | ||
KAFKA-9939 | Fetch request metrics are updated twice for delayed fetches | ||
KAFKA-9942 | ConfigCommand fails to set client quotas for default users with --bootstrap-server. | ||
オプションの内容／値の変更／省略値の変更（注2） | KAFKA-8855 | Collect and Expose Client's Name and Version in the Brokers | |
KAFKA-9040 | Implement --all option for describing configs | ||
KAFKA-9494 | Include data type of the config in ConfigEntry | ||
KAFKA-9612 | CLI Dynamic Configuration with file input | ||
チェック強化（注3） | KAFKA-9102 | Increase default zk session timeout and max lag | |
KAFKA-9324 | Drop support for Scala 2.11 (KIP-531) | ||
KAFKA-9507 | AdminClient should check for missing committed offsets | ||
KAFKA-9677 | Low consume bandwidth quota may cause consumer not being able to fetch data | ||
公開しているファイルの内容／形式（注4） | KAFKA-10134 | High CPU issue during rebalance in Kafka consumer after upgrading to 2.5 | |
KAFKA-9267 | ZkSecurityMigrator should not create /controller node | ||
KAFKA-9437 | KIP-559: Make the Kafka Protocol Friendlier with L7 Proxies | ||
KAFKA-9606 | Document Metrics Changes from KIP-444 | ||
KAFKA-9724 | Consumer wrongly ignores fetched records "since it no longer has valid position" | ||
KAFKA-9801 | Static member could get empty assignment unexpectedly | ||
KAFKA-10113 | LogTruncationException sets fetch offsets incorrectly | ||
KAFKA-10224 | The license term about jersey is not correct | ||
KAFKA-8768 | Replace DeleteRecords request/response with automated protocol | ||
KAFKA-8928 | Java producer default max message size does not align with broker default | ||
KAFKA-9047 | AdminClient group operations may not respect backoff | ||
KAFKA-9330 | Calling AdminClient.close in the AdminClient completion callback causes deadlock | ||
KAFKA-9364 | Fix misleading consumer logs on throttling | ||
KAFKA-9435 | Replace DescribeLogDirs request/response with automated protocol | ||
KAFKA-9539 | Add leader epoch in StopReplicaRequest | ||
KAFKA-9577 | Client encountering SASL_HANDSHAKE protocol version errors on 2.5 / trunk | ||
KAFKA-9605 | EOS Producer could throw illegal state if trying to complete a failed batch after fatal error | ||
KAFKA-9815 | Don't log passwords for AlterConfigs requests in request logs | ||
KAFKA-9863 | update the deprecated --zookeeper option in the documentation into --bootstrap-server | ||
KAFKA-9922 | Update examples README | ||
KAFKA-9987 | Improve sticky partition assignor algorithm | ||
メッセージ内容の変更（注5） | KAFKA-9025 | ZkSecurityMigrator not working with zookeeper chroot | |
KAFKA-9460 | Enable TLSv1.2 by default and disable all others protocol versions | ||
KAFKA-9477 | Doc: Add RoundRobinAssignor as an option to consumer configs | ||
KAFKA-9499 | Improve deletion process by batching more aggressively | ||
KAFKA-9750 | Flaky test kafka.server.ReplicaManagerTest.testFencedErrorCausedByBecomeLeader | ||
KAFKA-9752 | Consumer rebalance can be stuck after new member timeout with old JoinGroup version | ||
KAFKA-9826 | Log cleaning repeatedly picks same segment with no effect when first dirty offset is past start of active segment | ||
KAFKA-10212 | Describing a topic with the TopicCommand fails if unauthorised to use ListPartitionReassignments API | ||
KAFKA-10223 | ReplicaNotAvailableException must be retriable to handle reassignments | ||
KAFKA-4203 | Java producer default max message size does not align with broker default | ||
KAFKA-6144 | Allow serving interactive queries from in-sync Standbys | ||
KAFKA-6266 | Kafka 1.0.0 : Repeated occurrence of WARN Resetting first dirty offset of __consumer_offsets-xx to log start offset 203569 since the checkpointed offset 120955 is invalid. (kafka.log.LogCleanerManager$) | ||
KAFKA-8470 | State change logs should not be in TRACE level | ||
KAFKA-8507 | Support --bootstrap-server in all command line tools | ||
KAFKA-8834 | Distinguish URPs caused by reassignment plus other metrics | ||
KAFKA-8843 | Zookeeper migration tool support for TLS | ||
KAFKA-9016 | Warn when log dir stopped serving replicas | ||
KAFKA-9026 | Replace DescribeAcls request/response with automated protocol | ||
KAFKA-9026 | Consumer should handle `CORRUPT_MESSAGE` error code in fetch response | ||
KAFKA-9027 | Replace CreateAcls request/response with automated protocol | ||
KAFKA-9074 | Connect's Values class does not parse time or timestamp values from string literals | ||
KAFKA-9218 | MirrorMaker 2 can fail to create topics | ||
KAFKA-9307 | Transaction coordinator could be left in unknown state after ZK session timeout | ||
KAFKA-9313 | Set default for client.dns.lookup to use_all_dns_ips | ||
KAFKA-9316 | ConsoleProducer help info not expose default properties | ||
KAFKA-9320 | Enable TLSv1.3 by default and disable some of the older protocols | ||
KAFKA-9353 | Print groupInstanceId in Consumer group command | ||
KAFKA-9491 | Fast election during reassignment can lead to replica fetcher failures | ||
KAFKA-9519 | Deprecate ZooKeeper access for kafka-configs.sh | ||
KAFKA-9540 | Application getting "Could not find the standby task 0_4 while closing it" error | ||
KAFKA-9553 | Transaction state loading metric does not count total loading time | ||
KAFKA-9583 | OffsetsForLeaderEpoch requests are sometimes not sent to the leader of partition | ||
KAFKA-9601 | Workers log raw connector configs, including values | ||
KAFKA-9652 | Throttle time metric needs to be updated for KIP-219 | ||
KAFKA-9654 | ReplicaAlterLogDirsThread can't be created again if the previous ReplicaAlterLogDirsThreadmeet encounters leader epoch error | ||
KAFKA-9701 | Consumer could catch InconsistentGroupProtocolException during rebalance | ||
KAFKA-9716 | Values of compression-rate and compression-rate-avg are misleading | ||
KAFKA-9775 | IllegalFormatConversionException from kafka-consumer-perf-test.sh | ||
KAFKA-9777 | Purgatory locking bug can lead to hanging transaction | ||
KAFKA-9796 | Broker shutdown could be stuck forever under certain conditions | ||
KAFKA-9802 | Fix flaky system test `TransactionsTest.test_transactions` | ||
KAFKA-9820 | validateMessagesAndAssignOffsetsCompressed allocates batch iterator which is not used | ||
KAFKA-9823 | Consumer should check equality of the generation for coordinator requests | ||
KAFKA-9840 | Consumer should not use OffsetForLeaderEpoch without current epoch validation | ||
KAFKA-9952 | Reconsider immediate propagation of hw when replica selector is used | ||
KAFKA-9984 | Should fail the subscription when pattern is empty | ||
メッセージの追加・削除（注6） | KAFKA-10089 | The stale ssl engine factory is not closed after reconfigure | |
KAFKA-7251 | Add support for TLS 1.3 | ||
KAFKA-7639 | Incremental Rebalance Protocol for Kafka Consumer | ||
KAFKA-8988 | Replace CreatePartitions request/response with automated protocol | ||
KAFKA-9024 | org.apache.kafka.connect.transforms.ValueToKey throws NPE | ||
KAFKA-9051 | Source task source offset reads can block graceful shutdown | ||
KAFKA-9110 | Improve efficiency of disk reads when TLS is enabled | ||
KAFKA-9130 | Allow listing consumer groups per state | ||
KAFKA-9144 | Early expiration of producer state can cause coordinator epoch to regress | ||
KAFKA-9146 | Add option to force delete members in stream reset tool | ||
KAFKA-9254 | Metadata not updated when consumer encounters FENCED_LEADER_EPOCH | ||
KAFKA-9293 | NPE in DumpLogSegments with --offsets-decoder | ||
KAFKA-9373 | Improve shutdown performance via lazy accessing the offset and time indices. | ||
KAFKA-9420 | Bump APIs to enable flexible versions | ||
KAFKA-9695 | AdminClient allows null topic configs, but broker throws NPE | ||
KAFKA-9854 | Re-authenticating causes mismatched parse of response | ||
KAFKA-9864 | Avoid expensive QuotaViolationException usage | ||
KAFKA-9918 | SslEngineFactory is NOT closed when channel is closing | ||
KAFKA-9945 | TopicCommand should support --if-exists and --if-not-exists when --bootstrap-server is used | ||
KAFKA-9985 | Sink connector consuming DLQ topic may exhaust broker | ||
実行結果の変更 | 誤った実装の修正（注7） | KAFKA-10043 | Some parameters will be overwritten which was configured in consumer.config where running "ConsumerPerformance.scala" |
KAFKA-10268 | dynamic config like "--delete-config log.retention.ms" doesn't work | ||
KAFKA-8764 | LogCleanerManager endless loop while compacting/cleaning segments | ||
KAFKA-9232 | Coordinator new member heartbeat completion does not work for JoinGroup v3 | ||
KAFKA-9254 | Updating Kafka Broker configuration dynamically twice reverts log configuration to default | ||
KAFKA-9306 | Kafka Consumer does not clean up all metrics after shutdown | ||
KAFKA-9492 | ProducerResponse with record-level errors throw NPE with older client version | ||
KAFKA-9498 | Topic validation during the creation trigger unnecessary TopicChange events | ||
KAFKA-9658 | Removing default user quota doesn't take effect until broker restart | ||
KAFKA-9661 | Config synonyms are no longer included in kafka-configs --describe output | ||
KAFKA-9835 | Race condition with concurrent write allows reads above high watermark |
注1）実行結果、実行権限、実行多重度の変更など
注2）設定画面、操作画面など、画面情報含む
注3）指定可能範囲の変更、定義間の整合チェック、チェックの厳密化による有効範囲の拡大／縮小
注4）ログファイルの出力項目や形式の変更など
注5）ポップアップメッセージなどの変更により従前の操作が変わるものを含む。メッセージ内容、メッセージレベルの変更、メッセージ改善
注6）既存機能を使用する範囲で障害修正、改善などによるメッセージ新規追加・削除。
注7）外部仕様に反した外部動作を正規の動作に修正する場合、または誤った解釈の基で実装した標準的な技術の動作を正規の動作に修正する場合
参照
詳細は、下記のサイトを参照してください。
