I’ve an software with about 200ok customers and am operating a NGINX + Gunicorn(Python) server behind an AWS EC2 loadbalancer.
I do not perceive how my requests is all the time 4k/minute however solely generally I get half of site visitors being timeout points. More often than not all requests are fantastic, however generally it begins to lock up then virtually all requests get timeouts.
I observed this sample of the # of present connections has a wave and fluctuates from 1000s to 0. Is NGINX bundling requests in some way? How can I differentiate request_time to determine whether or not it is NGINX not being configured correctly, or my Python server simply getting further sluggish endpoints being referred to as too usually.
I’ve connected a screenshot of one of many servers in my NGINX Amplify dashboard.
Any concepts of components of NGINX logs or Amplify that I can examine to find out whether or not that is an NGINX configuration difficulty, or if the hosted Python course of is getting locked up? Thanks!