mjsax commented on PR #15449: URL: https://github.com/apache/kafka/pull/15449#issuecomment-1972488983
We have two runs of this test, one w/o error injection and one w/ error injection. The runs w/o error injections are more stable, and I did see a "verifyCommitted" that did see fail because of too much data (for keys/tasks that should not have committed), hence, I concluded that `commit.interval.ms` must have hit. This will be hard to verify on the test w/o error injection because it's pretty stable. I don't know right now, if it help for the error injection case. In fact, it could make it worse (I did not dig deep enough, but the test might actually require low TX-timeout...) -- if this is the case, we would need "hack" into `StreamsConfig` to allow us to disable the commit-interval/tx-timeout check so we can set a large commit interval and a small tx-timeout. I could not make the test fail locally, so I am hoping that a couple of Jenkins PR runs will give us some signal about how to move forward. -- 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: jira-unsubscr...@kafka.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
