starrysxy commented on issue #16285: URL: https://github.com/apache/dolphinscheduler/issues/16285#issuecomment-2290629944
> > @SbloodyS When I try to fix this bug, using the dev branch, I found it has been fixed in #15795. But the bug is still there in version 3.1.x. I'm not sure if I should commit to fix this bug in version 3.1.x. If yes, which brach should I use to commit a PR. > > After take a deep look at it, I found the task instance status still bind with workflow instance status in the `dev` branch. You mean current fix is not good enough, right? BTW, I found a new scenario which could cause the same problem. Both workflows are the same as https://github.com/apache/dolphinscheduler/issues/16285#issuecomment-2216171367 I set the failure retry time for A-3 to 5 minutes. Assume that A workflow failed at 7:00 am, because A-3 is failed. Then B workflow failed at 7:03, because at that time, A-3's status is failure. At 7:05, A-3 retried and succeeded. So, when I check the result of the both workflows, I found the status of A is success and B is failure. So, I think if a task like A-3 is depended by other tasks and a retry policy has been set, the status of the downstream task should be waiting until the upstream task has no more retries. -- 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]
