Nginx is a fast and light HTTP server, reverse proxy, load balancer (and more).
It’s pretty simple to get TurboGears set up behind a Nginx server so that it proxies requests to the CherryPy server. Here is a sample configuration that not only proxies to your TurboGears application, but serves static content with Nginx and load balances between two TurboGears application instances as well.
http {
# boilerplate nginx config ...
upstream mycluster {
server 127.0.0.1:8000;
server 127.0.0.1:8001;
}
server {
listen 80;
# static files
location ^~ /static/ {
root /path/to/YourProject/package;
}
location = /favicon.ico {
root /path/to/YourProject/package/static/images;
}
# proxy to turbogears app
location / {
proxy_pass http://mycluster;
proxy_redirect off;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
}
}
}
Next you need to setup a couple TurboGears backends that will comprise the cluster:
Edit your production configuration (e.g. prod.cfg) and add/change the following lines:
server.socket_host="127.0.0.1"
server.socket_port=8000
Copy prod.cfg to prod2.cfg and change the server.socket_port option:
server.socket_host="127.0.0.1"
server.socket_port=8001
Start both instances of your app:
$ python start-myproject.py prod.cfg &
$ python start-myproject.py prod2.cfg &
That’s it! Nginx should now be passing requests across both backends transparently.
You can find more information and recipes for setting up Nginx on the Nginx wiki.