Nginx via FastCGI

The first step in FastCGI integration, is to create the application server that Nginx will sent requests to. For this example, we’ll use a unix domain socket, rather than TCP/IP for connectivity between the two.

FastCGI process

Create myapp/fastcgi.py:

try:
    from flup.server.fcgi_fork import WSGIServer
except ImportError:
    from chula.vendor.fcgi import WSGIServer
    print "Unable to import flup.server.fcgi import WSGIServer"
    print " >>> Falling back on old version available in Chula"

from chula.www.adapters.fcgi import adapter

# Expose the myapp, as it's not "installed"
sys.path.insert(0, os.getcwd())

from model import configuration

@adapter.fcgi
def application():
    return configuration.app

# Start the server which will handle calls from the webserver
WSGIServer(application, bindAddress='/tmp/myapp.socket').run()

Start up the FastCGI process:

python myapp/fastcgi.py

Make sure Nginx has permissions to write to the socket:

chmod o+w /tmp/myapp.socket

TODO: Provide an example init script to properly startup the socket, setting permissions and what not.

Nginx config

Configure Nginx to proxy application requests to our application. Add this to the server block of your Nginx configuration:

# Send all requests without a file extension to myapp:
location ~ ^([a-z/_])+$ {
  # This is needed when using Ubuntu
  include /etc/nginx/fastcgi_params;

  # FastCGI parameter settings
  fastcgi_read_timeout 3m;
  fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
  fastcgi_param SERVER_ADMIN NA;
  fastcgi_param SERVER_SIGNATURE nginx/$nginx_version;

  # The path to our running unix domain socket
  server unix:/tmp/myapp.socket;
}

Restart Nginx

sudo /etc/init.d/nginx restart

Try it!

Now you should be able to hit: http://your-server/myapp/blog