steveloughran commented on PR #6938: URL: https://github.com/apache/hadoop/pull/6938#issuecomment-2326391420
this is funny: placing the simple PUT in the same thread as close() breaks `ITestS3AConcurrentOps`. Why so? that test looks for at least one thread called s3a-transfer then asserts that after the thread timeout that count goes to zero. It is meant to assert that after renames the pool is drained but we've made two changes this year to reduce the #of threads 1. small file renames don't use copy manager. This invalidated the test -we just never noticed 2. this PR: small PUT doesn't do async. this caused the regression caused by change 1 to surfacen As a result: no threads to assert on. I'm fixing it by shrinking the size of multipart uploads to their minimum -this *seems* to work, though if problems surface in future we should look at the test and decide whether or not it is obsolete -or whether we could redesign the tests to include more parallelized operations (tree renames?) -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
