SEZ9 commented on PR #11382:
URL: https://github.com/apache/seatunnel/pull/11382#issuecomment-5643196027

   Thanks for the per-finding status at `1ee22aced026` — building on that 
rather than restarting:
   
   - **F1** — Agreed. `checkpointCompleted` being `synchronized` with 
`notifyAll()` on the qualifying branch, paired with the `synchronized` wait 
loop in `checkSchemaChangeCanApply()`, matches the intended fix shape. Treating 
F1 as resolved at this head.
   - **F2** — Keeping open, as noted. The runtime mechanism 
(`restoreCoordinator(false)` resetting `isAllTaskReady`) is correct and pinned 
by the coordinator test, but the `checkpointCompleted` javadoc still carries 
the sentence claiming the coordinator is always recreated on task restart. 
Please push a javadoc update describing the actual re-arming mechanism before 
merge.
   - **F3 / F8** — The write-side fail-fast with `IOException` on a null 
`checkpointType` is noted. Still open: read-side handling of an unknown type 
name in `CheckpointFinishedOperation`, and whether a test covers round-trip 
serialization including those cases.
   - **F4** — Is drain-ready state still process-memory only, with notification 
re-delivery as the sole recovery path? If so, please state that explicitly in 
the guard's javadoc.
   - **F5** — Is there now a coordinator-side call site that sends a typed 
abort, or should the guard's typed-abort cleanup branch be removed or 
documented as unreachable?
   - **F6** — How are overlapping schema changes handled with the single-slot 
tracking, and is the type-based abort reset still broader than the checkpoint 
it targets?
   - **F7** — Beyond the guard unit tests and the coordinator test, is there 
coverage of the `SinkFlowLifeCycle` / `SeaTunnelTask` wiring or an end-to-end 
schema-evolution-with-failover scenario?
   
   Happy to re-review once the F2 javadoc fix and the remaining items land. 
Thanks!
   
   <!-- streview-comment:992 -->


-- 
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]

Reply via email to