DanielLeens commented on PR #12258: URL: https://github.com/apache/seatunnel/pull/12258#issuecomment-5664679605
Pushed `10a35041ef` addressing the review's Issue 1: `query(String)` wraps any `SQLException` as a bare `RuntimeException`, which `Awaitility.untilAsserted` does not retry, so it would propagate immediately out of `await()` and past the `ConditionTimeoutException`-only `catch` from `16e377bb7b`, silently skipping the dump on that path. Given this test drops/re-adds a primary key on the sink table right around the restore it's exercising, a transient `SQLException` during that window is plausible, not hypothetical. Added a second `catch (RuntimeException e)` that also fires the dump when `e.getCause() instanceof SQLException`, and pulled the dump itself into a `dumpTableContentsOnBestEffort` helper that swallows a failure in the dump's own queries (logs a warning) rather than let it mask the original failure being explained. Also switched the fully-qualified `ConditionTimeoutException` reference to a normal import per the earlier nit. Still test-only, no production code touched. This doesn't change the open question (Issue 2, the actual source/sink divergence) — it just makes sure the next CI run's diagnostic actually fires regardless of which of the two ways this test can fail. -- 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]
