Asquator commented on PR #53492: URL: https://github.com/apache/airflow/pull/53492#issuecomment-3168927238
@dstandish > This query is not a starvation panacea, I don't think. > > For example what if there are two conditions interplaying. > > Suppose there's only one pool slot available in pool A. Tasks X and Y have pool A. Task X sorts first. So task Y would be filtered out by the pool constraint. > > But suppose further that for task X, total_tis_per_dagrun_count is beyond dag max active tasks. So nothing will be scheduled for pool A even though Y should be scheduled. I think this is a very likely situation, and I think this was the reason I decided to nest the windows in the beginning. I'll have to think it over. If we nest the windows, the problem doesn't exist anymore since every window gets the filtered TI.ids. -- 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]
