SEZ9 commented on PR #12081: URL: https://github.com/apache/seatunnel/pull/12081#issuecomment-5642807260
@Rangsh thanks for the detailed follow-up on `f17373ac5`. **1. Survivability test** — the assertions you describe in `WALWorkHandlerSurvivabilityTest.nonIoExceptionFromWriteShouldNotKillWorkerAndSubsequentAppendIsFailClosed` are exactly the behavior I was looking for on F1/F2/F4: worker survives a non-`IOException` from `writer.write()`, the first future completes with `done(false)`, and the second APPEND is fail-closed with `writer.write()` invoked only once and `isAppendBlockedAfterWriteFailure()` sticky. I'll verify that against the code on `f17373ac5` before marking F4 resolved. **2. Fail-close on flush/sync** — your reasoning (`catch (Exception)` around `writer.write(...)` plus `HdfsWriter.write` always ending in `flush()`) makes sense. Your message appears to be cut off at "complete with `done(false)` wit" — if there was a caveat after that (for example a path where the flag is *not* set), could you post the rest? If there is already a test that drives the failure through `flush()` throwing rather than `write()` throwing, please point me to it; otherwise I'd suggest adding one so the fail-close-on-sync behavior is pinned by a test rather than by the current call structure. <!-- streview-comment:975 --> -- 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]
