Important
You are viewing documentation for an older version of Confluent Platform. For the latest, click here.
Changelog¶
Version 3.3.0¶
Version 3.2.2¶
Version 3.2.1¶
No changes
Version 3.2.0¶
Version 3.1.1¶
No changes
Version 3.1.0¶
- PR-134 - Flush the last partial file when incoming stream is paused.
- PR-133 - CC-331: Update config options docs
- PR-126 - Fix TimeBasedPartitioner config validation
- PR-112 - Lint change to avoid compiler error in Oracle JDK 1.7 using jenv.
- PR-94 - Fix lint annoyances
- PR-108 - Revert “support multi partition fields.”
- PR-105 - support multi partition fields.
- PR-101 - Added link to Confluent documentation for the connector.
- PR-92 - Start a new WAL file after truncate instead of appending to log.1
- PR-87 - Scheduled rotation implementation
- PR-90 - Use configured Hadoop configuration object for Parquet writer
- PR-91 - Upgrade to Hadoop 2.6.1
- PR-70 - Fix handling of topics with periods
- PR-68 - prints details of HDFS exceptions
- PR-67 - clean up hive metastore artifacts from testing
- PR-64 - cleaned up .gitignore. Now ignores Eclipse files
Version 3.0.0¶
HDFS Connector¶
- PR-62 - Update doc for CP 3.0.
- PR-60 - Remove HDFS connectivity check.
- PR-55 - Removing retry logic from HiveMetaStore to fix the metastore connection bloat.
- PR-50 - Remove close of topic partition writers in DataWriter close.
- PR-42 - Using new config validation.
- PR-41 - Bump version to 3.0.0-SNAPSHOT and Kafka dependency to 0.10.0.0-SNAPSHOT.
- PR-35 - Minor doc typo fix TimeBasedPartitioner.
- PR-33 - Minor doc fix.