SEZ9 commented on issue #10203: URL: https://github.com/apache/seatunnel/issues/10203#issuecomment-6378553430
Classification: D / Connector-V2 MySQL-CDC live table discovery and exactly-once recovery design. @ryanmeowy, one short follow-up on the transition table; everything else in my previous comment (design-only hold, acceptance contract, test plan) stands as written. Row 2b: the table says the split is handed back but T's snapshot does not re-run. When the backfill has finished but the checkpoint carrying `W_high` never completed, which fields in the reader's checkpointed split state tell the restored reader that the split is complete rather than pending? Please spell that out in the design revision, and confirm that the `shouldEmit()` comparison in row 3 is against `tableWatermarks[T]` per table rather than a global watermark. Thanks for laying the cuts out this clearly. <!-- streview-comment:1980 --> -- 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]
