We are using uwsgi to serve the python app behind nginx.
Can I hope that the theoretical maximum of simultaneously served client connections is equal to uwsgi processes * threads?
We are using server-sent events, so, the connections are held for a long time.