I would like to set a duration for Nginx to terminate the workers after the timeout. I am looking for a feature like "Ping Maximum Response Time" in IIS. I am using uWSGI, Nginx, and Pylons Framework.
IIS - Ping Maximum Response Time (seconds): Maximum time (in seconds) that a worker process is given to respond to a health monitoring ping. If the worker process does not respond, it is terminated.
I found some Nginx directives, but I am not sure which one(s) must be used.
I found the solution.
As Nginx is used as a reverse proxy, I added these three directives.
proxy_connect_timeout 90;
proxy_send_timeout 90;
proxy_read_timeout 600; # (in seconds) <- must be changed to what we need.