[Bug] data loss when enable 2pc #205
Description
Search before asking
- I had searched in the issues and found no similar issues.
Version
1.3.1
What's Wrong?
The data is lost without any exception when enable doris.sink.enable-2pc and spark jobs running very long time e.g. 3~4hour.
Spark job driver warn logs.
The transcation is removed bacause of reaching timeout limit.
What You Expected?
It should throw exception when failed to commit transaction .
How to Reproduce?
No response
Anything Else?
No response
Are you willing to submit PR?
- Yes I am willing to submit a PR!
Code of Conduct
- I agree to follow this project's Code of Conduct