SEZ9 commented on PR #12081: URL: https://github.com/apache/seatunnel/pull/12081#issuecomment-6552868383
Thanks for the follow-up on `b172d96d`. Going back over the earlier points from `7f1f5b09`: - **F6 (sequential per-entry waits in batch)** — a single shared deadline per batch in `IMapFileStorage` is the shape I was hoping for. Could you point me to the hunk so I can confirm it before closing this one? - **F4 (writer reused after a failed append)** — fail-closing the writer after the first append failure and answering later events as failed without touching the stream would address the torn-record risk. Same request: please point me to the `WALWorkHandler` change. - **F2 (worker dying on `executeResponse()` throwing)** — please confirm the new exception handling in `WALWorkHandler` also wraps the response/completion call, not only the write itself. - **F3 (semantic change of `get`)** — separating completion from success and surfacing `TimeoutException` from the timed `get(remaining, unit)` is fine provided all callers in this module were updated together; please confirm. Still open before I approve: 1. **F1** — what does the untimed `RequestFuture.get()` do now? If it is still an unbounded block, either remove it or cap it with `writDataTimeoutMilliseconds`, so a worker killed by an `Error` or a never-dispatched event cannot hang a caller forever. 2. **F7** — please add method-level Javadoc on `get(timeout, unit)` stating the throw-on-timeout contract, not just the class-level note. 3. **F5** — `HdfsWriterFlushSyncPathTest` uses Mockito, but I do not see a `pom.xml` change adding the test dependency to `imap-storage-file`. Either point me to where it is inherited from, or add it explicitly. 4. **F8** — a timed-out write wait in `queryExecuteStatus` is an expected condition; please log it at WARN with a one-line message (future id + timeout) rather than a full `TimeoutException` stack trace at ERROR, which will flood logs under a stuck worker. Once those are in, I will do a final pass on the head. <!-- streview-comment:2190 --> -- 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]
