GitHub user hanxdatadog added a comment to the discussion: Best practice for 
webserver liveness probe check

> For whatever reason this wasn't an issue on airflow 2.x but on 3.x I've had 
> to increase the number of connections I use to the db server

We also had to increase the connection pool size. Thought it was expected to do 
so. glad to know it wasn't an issue in airflow 2. From what I could see, after 
deployment everything would be healthy for a while but at some point the 
connection pool would get consumed up and the webserver performance was 
severely downgraded. But after we bumped the connection pool size to something 
higher, never seen that problem anymore. seems like we do have a minimal 
connection pool size requirement that implicitly derived from some parallelism 
configurations in the scheduler. (a.k.a if you want X parallelism, db 
connection pool size must be > f(X))

GitHub link: 
https://github.com/apache/airflow/discussions/54853#discussioncomment-15063210

----
This is an automatically sent email for [email protected].
To unsubscribe, please send an email to: [email protected]

Reply via email to