Hi Bolang, I'm not sure what you meant by "simply using nginx" since as far as I know it doesn't have standard WSGI module. I hope you're not talking about running WSGI application as CGI script, do you?
As to comparing Gunicorn to gevent-fastcgi Gunocorn is HTTP server and gevent-fastcg is FastCGI one. FastCGI can run over UNIX domain sockets, it can multiplex requests using single connection. That could be deal-breaker when application is using long-polling requests and there are thousands of clients that can be accessing it simultaneously. Beside HTTP is not well suited for communication between frontend and backend (just recall famous need to fix hostname:port in backend server responses because backend server has no way to get access to original HTTP-request and it might not even know that there is something that makes request on behalf of client browser). There is no such problems in FastCGI protocol because it was specifically designed for frontend/backend communication. Gevent-fastcgi is not well suited for any type of applications. The application should not block since everything is run in the same thread (there are greenlets that are used in place of thread to avoid GIL). Not all databases and other external resources can be used in non-blocking mode. Luckily PostgreSQL, MySQL, Memcache are those that have adapters/hooks for making them work well in non-blocking manner. Alex Bolang wrote: > Hi Alexander, > What is the advantage of using gevent-fastcgi instead of simply using > nginx and maybe with gunicorn? -- You received this message because you are subscribed to the Google Groups "Django users" group. To post to this group, send email to django-users@googlegroups.com. To unsubscribe from this group, send email to django-users+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/django-users?hl=en.
