Release Note 1.5.0

[JNSimba]

Feature & Improvement

1. Support Doris’ Map, struct type #212
2. Optimize schema change #167 #205
3. Support direct connected be writing #187
4. Supports batch writing and does not rely on checkpoint #168
5. Can be directly connected to FE for writing, and uniq table is closed by default 2pc #202
6. Support oracle, sqlserver, postgresql whole database synchronization #181 #178
7. Optimize the synchronization part of the entire database: synchronize mysql tables without primary keys, add uid, upgrade cdc version to 2.4.2, etc.
8. Introduce CacheRecorebuffer #193
9. Support capturing update-before event #169
10. Support table creation through flink catalog #147
11. sink api reconstruction #213
12. DorisSink, DorisBatchSink support multi-table writing #224 #223
13. The entire database synchronization supports single Sink writing #245
14. Whole database synchronization supports automatic table creation #248

Bug

1. Fix the problem of data loss during two-stage commit #210
2. The problem of duplicate data in asynchronous lookup query #198
3. Fix the problem of largetint reading error in doris2.0 #173
4. Fix the issue where resources are not released during lookup restart #162
5. Fix the time zone problem encountered during mysql whole database synchronization #185
6. Fix the problem of integer overflow in catalog creation table

Thanks

Thanks to everyone who has contributed to this release: @bingquanzhao @cjj2010 @codeAntg @cygnusdark @DongLiang-0 @GoGoWen @huangkai66 @IKAZUCHI-Akira @JNSimba @scxwhite @SunShine-1204 @thehuldra @vinlee19 @wanshicheng @wolfboys @zy-kkk