potiuk commented on PR #29406: URL: https://github.com/apache/airflow/pull/29406#issuecomment-1461703484
> If we define “fail” on the DAG level as “a task will not triggered when all its upstreams finish” I think it can work logically; the question is whether this is expected by users (and we probably should invent a term other than “fail” to avoid confusion). Maybe "cancel" or "stop" ? Yes I think this is a useful cases in a number of scenarios especially when users are cost or timing (or both) conscious. There might be a number of cases, especially when you have dynamic cloud infrastructure, and long running tasks, where you know that failure of any task will generally make any other task results useless - so running them would be a waste of money (or time waiting for things to complete). -- 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]
