potiuk commented on PR #60166: URL: https://github.com/apache/airflow/pull/60166#issuecomment-3718936836
I also do not thik it's worth to implement different logic here - it always complicates things, and in this case I think performance loss is negligible. Logically speaking it's the same operation and while yes it is one more fetch, I am not sure it has a lot of impact on performance. The only potential issue that might happen is that the list of stale dag_ids is really long and takes a lot of memory (and maybe even exceeds the size of sql query generated in the second step) - but I doubt you might have big number of those - as this query is run **directly** after a loop that deactivates dags. Do you really tink it's worth to split the logic here? I am not even sure if in this particular case this is purely MySQL induced (it's likely - and happened in the past but we have not proven it yet) - so "protectively" it would be better to split it for all databases I think. -- 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]
