How to configure nginx for JBoss?

How would I configure Nginx and Gunicorn to be optimized/scalable?

• Being nginx is a compiled and cpu-optimized binary whereas gunicorn uses greenlets and is bytecode, will having only one instance of gunicorn running be a bottleneck in a reverse proxy configuration? Architecturally speaking, nginx can handle more resources than gunicorn can during a reverse proxy as nginx will simply pass connections to the one running gunicorn instance. The bulk of the work will be within the WSGI app. What is the preferred method of optimization? Should I have two instances of gunicorn running and configure nginx to distribute the load in a round robin approach?

• Answer:

  To start with NGINX is not a threaded web server. It uses the asynchronous model.  The first thing to do is make sure you are sure what this means, how it helps with better using of system resources and then maybe refine your question. There is more than enough information on the www to discuss NGINX performance and how it architecturally stands up against other architectures.