hey, I'm getting server errors under heavy loads when self hosting on k8s

sqlalchemy.exc.TimeoutError: QueuePool limit of size 5 overflow 10 reached, connection timed out, timeout 30.00 (Background on this error at: <https://sqlalche.me/e/20/3o7r>)

this causes flows to crash and sometimes creates zombie flows that I have to cancel/delete maually the loads are around 100 flows (each flow has around 100 sub-tasks), and flows are running on Process type agents I can increase these settings on the server, or maybe even put in a better Postgres DB, but i'm not sure how to avoid these errors in the future

Hi @Lior Barak, did you ever resolve this?

hi @Benji a very late response: I changed my DB to work with an external AWS RDS changed the server sqlalchemy settings:

PREFECT_SQLALCHEMY_POOL_SIZE

PREFECT_SQLALCHEMY_MAX_OVERFLOW

(notice that it can have adverse affects on your interaction with the DB) also, I have multiple prefect servers running in parallel and a limited amount of agents so I won't get to this overflow

Awesome, I came to a similar solution. Thanks for the update!