wanglijie95 commented on PR #20097: URL: https://github.com/apache/flink/pull/20097#issuecomment-1321372510
Thanks @eskabetxe for creating this PR. I have 3 questions about the changes: 1. Do we need to introduce `JdbcProducer/NonXaJdbcProducer`? I think the existing `JdbcOutputFormat` can cover all functionality of non-xa sink, can we reuse it? 2. Does the non-xa sink need to implement `StatefulSink` and `TwoPhaseCommittingSink`? I think the non-xa sink neither has state nor supports 2PC. 3. I don't see support for xa sink, is it not in the current plan? Some my thoughts: I think we can introduce 2 implementations: `JdbcXaSink` and `JdbcNonXaSink`. `JdbcXaSink` implements `StatefulSink` and `TwoPhaseCommittingSink`, the `JdbcNonXaSink` implements the based `org.apache.flink.api.connector.sink2.Sink`. These two implementations can then be exposed to the user via `.sink` and `.exactlyOnceSink`, just like the legacy sink in `org.apache.flink.connector.jdbc.JdbcSink`. WDYT? -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
