Releases: pingcap/tiflow
Releases · pingcap/tiflow
ticdc v4.0.5
New Features
- Support Kafka SSL connection #764
- Support outputting the old value #708
- Add the column flags #796
- Support outputting the DDL statements and table schema of the previous version #799
Improvements
- Lower the frequency of getting TSO #801
Bug Fixes
- Fix the issue that the failed
changefeed
cannot be removed #782 - Fix invalid
delete
events by selecting one unique index as the handle index #787 - Fix the bug that GC safepoint is forwarded beyond the checkpoint of stopped
changefeed
#797 - Fix the bug that the network I/O waiting blocks tasks to exit #825
ticdc v4.0.4
Note: The current version of TiCDC is still experimental and not recommended for use in a production environment.
ticdc v4.0.3
Note: The current version of TiCDC is still experimental and not recommended for use in a production environment.
ticdc v4.0.2
Note: The current version of TiCDC is still experimental and not recommended for use in a production environment.
ticdc v4.0.1
Note: The current version of TiCDC is still experimental and not recommended for use in a production environment.
ticdc v4.0.0
Note: The current version of TiCDC is still experimental and not recommended for use in a production environment.
Changed Behaviors
- Optimized the structure of Changefeed configuration file #588
- Filter transactions by the start ts instead of the commit ts #589
New Features
- Check the validity of Sink URI when creating the Changefeed #561
- Check the PD and TiKV versions when starting TiCDC Server #570
- Optimized the table scheduling mechanism to support scheduling multiple tables in one tick period #572
- Add the node role information in HTTP API #591
Bug Fix
ticdc v4.0.0-beta.2
Initial release of the change data capture, providing following features in this release
- Support capturing change data from TiKV since v4.0.0-beta.2
- Support replicating change data to MySQL protocol compatible database, with a eventual consistency guarantee
- Support replicating change data to Kafka, with either a eventual consistency guarantee or row level order guarantee
- Provide a native support for high availability