kafkajs

[1.9.0] - 2019-06-25

Added

• Add typescript declaration file #362 #385 #390
• Add requestTimeout to apiVersions #369
• Discard messages saw a seek operation during a fetch or batch processing #367
• Include fetched offset metadata retrieved with admin.fetchOffsets #389
• Allow offset metadata to be written as part of OffsetCommit requests #392
• Prevent the consumption of messages for topics paused while fetch is in-flight #397
• Add AWS-IAM SASL mechanism #402
• Add batch.offsetLagLow #405

Changed

• Don't modify global.crypto #365
• Change log level about producer without metadata #382
• Update encoder to write arrays as single Buffer.concat where possible #394

Fixed

• Log error message on connection errors #400
• Make sure runner has connected brokers and fresh metadata before it starts #404

[1.8.0] - 2019-05-13

Added

• Add partition-aware concurrent mode for eachMessage #332
• Add JavaCompatiblePartitioner #358
• Add consumer.subscribe({ topic: RegExp }) #346
• Update supported protocols to latest of Kafka 1 #343 #347 #348

Changed

• Add documentation link to REBALANCE_IN_PROGRESS error #341

Fixed

• Fix crash on offline replicas in metadata v5 response #350

[1.0.0] - 2018-05-14

Changed

• Updated readme

[1.0.1] - 2018-05-18

Fixed

• Prevent crash when re-producing after metadata refresh #62

[1.1.0] - 2018-06-14

Added

• Support to SASL SCRAM (scram-sha-256 and scram-sha-512) #72
• Admin client with support to create topics #73

[1.3.0] - 2018-08-06

Fixed

• Skip unsubscribed topic assignment #86
• Refresh metadata when producing to a topic without metadata #87
• Discard messages with a lower offset than requested #100

Added

• Add consumer auto commit policies #89
• Notify user when setting heartbeat interval to same or higher than session timeout #91
• Constantly refresh metadata based on metadataMaxAge #94
• New instrumentation events #95
• Expose loggers #97 #102
• Add offset management operations to the admin client #101
• Support to record batch compression #103
• Handle missing username/password during authentication #104

[1.3.1] - 2018-08-20

Fixed

• Client logger accessor #106
• Producer v3 decode format #114
• Parsing multiple responses #115
• Fetch v4 for partial messages on record batch #116

Added

• Connection instrumentation events #110

[1.4.0] - 2018-10-09

Fixed

• Potential offset loss when updating offsets for resolved partitions #124
• Refresh metadata on lock timeout #131
• Cleans up stale brokers on metadata refresh #131
• Force metadata refresh on ECONNREFUSED #134
• Handle API version not supported #135
• Handle v0.10 messages on v0.11 Fetch API #143

Added

• Admin delete topics #117
• Update metadata api and allow to disable auto topic creation #118
• Use highest available API version #135 #146
• Admin describe and alter configs #138
• Validate message format in producer #142
• Consumers can detect that a topic was updated and force a rebalance #136

Changed

• Improved stack trace for KafkaJSNumberOfRetriesExceeded #123
• Enable Kafka v0.11 API by default #141 (Can still be disabled with allowExperimentalV011=false)
• Replace event emitter Lock #154
• Add member assignment to GROUP_JOIN instrumentation event #136

[1.4.1] - 2018-10-17

Fixed

• Decode multiple RecordBatch on protocol Fetch v4 #179
• Skip incomplete record batches #182
• Producer with acks=0 never resolve #181

Added

• Runtime flag for displaying buffers in debug output #176
• Add ZSTD to compression codecs and types #157
• Admin get topic metadata #174

Changed

• Add description to lock instances #178

[1.4.2] - 2018-10-22

Changed

• Allow messages with a value of null to support tombstones #185

[1.4.3] - 2018-10-22

Changed

• Version 1.4.2 without test files

[1.4.4] - 2018-10-29

Fixed

• Protocol produce v3 wasn't filtering undefined timestamps and was sending timestamp 0 (NaN converted) for all messages #188

[1.4.5] - 2018-11-28

Fixed

• Fix crash in mitigation for receiving metadata for unsubscribed topics #221

Added

• Add CRASH instrumentation event for the consumer #221

[1.4.6] - 2018-12-03

Fixed

• Always assign partitions based on subscribed topics #227

[1.4.7] - 2019-01-17

Fixed

• Rolling upgrade from 0.10 to 0.11 causes unknown magic byte errors #246

[1.4.8] - 2019-02-18

Fixed

• Handle undefined message key when producing with 0.11 #247
• Fix consumer restart on find coordinator errors #253
• Crash consumer on codec not implemented error #256
• Fix error message on invalid username or password #270

[1.5.0-beta.0] - 2019-01-08

Changed

• Add transactional attributes to record batch #199
• Ignore control records #208
• Filter aborted messages on the consumer #223 #210 #228
• Make Round robin assigner forward userdata #231

Added

• Protocol FindCoordinator v1 #189
• Protocol InitProducerId v0 #190
• Protocol AddPartitionsToTxn v0 #191
• Protocol AddOffsetsToTxn v0 #194
• Protocol TxnOffCommit v0 #195
• Protocol EndTxn v0 #198
• Protocol ListOffsets v1 and v2 #217 #209
• Accept max in-flight requests on the connection #216
• Idempotent producer #203
• Transactional producer #206
• Protocol SASLAuthenticate #229
• Add SendOffsets to consumer eachBatch #232
• Add network instrumentation events #233
• Allow users to provide offsets to the commitOffsetsIfNecessary #235

[1.5.0-beta.1] - 2019-01-17

Fixed

• Rolling upgrade from 0.10 to 0.11 causes unknown magic byte errors #246

Changed

• Validate consumer groupId #244

Added

• Expose network queue size event to consumers, producers and admin #245

[1.5.0-beta.2] - 2019-02-13

Fixed

• Handle undefined message key when producing with 0.11 #247
• Fix consumer restart on find coordinator errors #253
• Crash consumer on codec not implemented error #256
• Fix error message on invalid username or password #270
• Restart consumer on crashes due to retriable error #269
• Remove deleted topics from the cluster target group #273

Changed

• Change Node engine requirement to >=8.6.0 #250
• Don't include lockfile and vscode files in package #264

Added

• Allow configuring log level at runtime #278

[1.5.0-beta.3] - 2019-02-20

Fixed

• Missing default restart time on crashes due to retriable errors #283
• Add custom requestTimeout for JoinGroup v0 #293
• Fix propagation of custom retry configs #295

Changed

• Allow calling Producer.sendBatch with empty list #287
• Encode non-buffer key as string by default #291