li3zhi4 commented on PR #11677: URL: https://github.com/apache/seatunnel/pull/11677#issuecomment-5882570813
Status update and a merge-readiness checklist for this PR \u2014 it's been quiet for a while, so here's the current state to make it easy to pick up:\n\n**Change (unchanged, head `5ba2e3c518`):** after a checkpoint recovery, restored splits that were left assigned to non-alive readers are now reassigned to waiting readers instead of stalling \u2014 `IncrementalSourceEnumerator` uses a new `SourceSplitEnumerator#registerWaitingReader/addSplitsBack` path, with the corresponding `SourceSplitEnumeratorTask` handling on the engine side. Tests: `IncrementalSourceEnumeratorTest` extended, plus a MySQL CDC E2E recovery scenario (`mysqlcdc_to_mysql_with_sink_failure_recovery.conf`).\n\n**Freshness:** I've just synced the branch with the latest `upstream/dev` \u2014 conflict-free. This brings in two related CDC changes that have since landed upstream: `#11271` (prune removed tables from restored incremental splits) and `#11503` (restore runtime schema from checkpoint after failover). They'r e complementary to this fix \u2014 this PR covers reassignment to waiting readers, not table pruning or schema restore \u2014 and the merge shows no interaction. CI is running on the new head now; I'll post results here.\n\n**Ask:** could @davidzollo or @nzw921rx (or anyone with context on CDC split lifecycle) take a look and confirm whether anything else is needed for this to move forward \u2014 or let me know if the preferred direction now differs after `#11271`/`#11503`? Happy to adapt.\n -- 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]
