We do a lot of patch requests. We have a large database (>10Gb). Lots of idl transactions in PostgreSQL. Very long transactions due to which the service is not available (nginx sends a 504 error). Maybe you have settings for postgreskull or nginx. Or any tips on how we can speed things up.
Hello @MrRaccoon, apologies for not getting back to you earlier! Can you please describe your request by following this template?
Hi @MrRaccoon, would you mind sharing more information on the patch requests that are being made? If the same row is updated over and over again, it could result in a problem where the row is locked, while another request is being made, forcing the gunicorn worker to wait for the lock to be released. If this happens a lot, it could eventually take up all gunicorn workers, resulting in 504 errors.