how to nginx virtual servers + fcgi for django?

Question

I would like to run several virtual hosts via nginx, each serving a different django app via fcgi. Is this possible? If so, does anyone have good resources on where/how to start? The nginx docs seem to be mostly examples, but none of the particular config I'm attempting...

Answer 1

There have been two pretty good blog posts lately about setting up nginx, but both are using nginx to serve static (or cached) content and apache + mod_wsgi to serve Django. I prefer this myself because mod_wsgi makes process management so much easier than fcgi (among other advantages).

Anyway, here are the posts:

• The Django and Ubuntu Intrepid Almanac
• django, nginx, memcached - the dynamic trio

EDIT: ok I dug out an old tar file with my nginx + django + virtual host config files from a year ago before I switched to apache + mod_wsgi. This was a development machine, so you'll want to adjust # of workers, connections, etc.

nginx.conf:

user nginx nginx;
worker_processes 2;

error_log /var/log/nginx/error_log info;

events {
    worker_connections  1024;
    use epoll;
}

http {
    include         /etc/nginx/mime.types;
default_type    application/octet-stream;

log_format main
            '$remote_addr - $remote_user [$time_local] '
    '"$request" $status $bytes_sent '
            '"$http_referer" "$http_user_agent" '
    '"$gzip_ratio"';

client_header_timeout   3m;
client_body_timeout     3m;
send_timeout            3m;

connection_pool_size            256;
client_header_buffer_size       1k;
large_client_header_buffers     4 2k;
request_pool_size               4k;

gzip on;
gzip_min_length 1100;
gzip_buffers    4 32k;
gzip_types      text/plain text/html application/x-javascript text/xml text/css;

output_buffers  4 32k;
postpone_output 1460;

sendfile        on;
tcp_nopush      on;
tcp_nodelay     on;

keepalive_timeout       75 20;

ignore_invalid_headers  on;

include vhosts.d/*.conf;
}

So, the main nginx.conf includes every file in the vhosts.d/ subdirectory. One of my files was for serving PHP on localhost (was probably running phpMyAdmin) like this:

vhosts.d/00_localhost:

server {

  listen 127.0.0.1:80;
  server_name localhost;

  access_log /var/log/nginx/localhost.access_log main;
  error_log /var/log/nginx/localhost.error_log info;

  root /var/www/localhost/htdocs;
  index index.php index.html index.htm;

    location ~ .php$ {
        fastcgi_pass   127.0.0.1:8888;
        fastcgi_index  index.php;
        fastcgi_param  SCRIPT_FILENAME  /var/www/localhost/htdocs$fastcgi_script_name;
        fastcgi_param  QUERY_STRING     $query_string;
        fastcgi_param  REQUEST_METHOD   $request_method;
        fastcgi_param  CONTENT_TYPE     $content_type;
        fastcgi_param  CONTENT_LENGTH   $content_length;
  }
}

And then a local Django development server like this:

50_django.local:

server {

  listen  80;
  server_name django.local;

  access_log /var/log/nginx/django.access.log main;
  error_log /var/log/nginx/django.error.log info;

  root /var/www/django.local;

  location ~* ^.+\.(mpg|avi|mp3|swf|zip|tgz|gz|rar|bz2|doc|xls|exe|ppt|txt|tar|mid|midi|wav|rtf|mpeg)$ {
    access_log off;
    break;
  }

  location ~* ^.+\.(jpg|jpeg|gif|png|ico|css|bmp|js)$ {
    expires 30d;
    break;
  }

  location / {
    fastcgi_pass unix:/var/www/django.local/server.sock;
    include fastcgi.conf;
  }

  location ^~ /admin/ {
    fastcgi_pass unix:/var/www/django.local/server.sock;
    include fastcgi.conf;
    access_log off;
    auth_basic "Welcome to admin";
  }
}

Finally, each of the virtual servers included a fastcgi.conf for each location.

fastcgi.conf:

fastcgi_pass_header Authorization;
fastcgi_intercept_errors off;

fastcgi_param PATH_INFO         $fastcgi_script_name;
fastcgi_param REQUEST_METHOD    $request_method;
fastcgi_param QUERY_STRING      $query_string;
fastcgi_param CONTENT_TYPE      $content_type;
fastcgi_param CONTENT_LENGTH    $content_length;
fastcgi_param SERVER_PORT       $server_port;
fastcgi_param SERVER_PROTOCOL   $server_protocol;
fastcgi_param SERVER_NAME       $server_name;

fastcgi_param REQUEST_URI       $request_uri;
fastcgi_param DOCUMENT_URI      $document_uri;
fastcgi_param DOCUMENT_ROOT     $document_root;
fastcgi_param SERVER_ADDR       $server_addr;
fastcgi_param REMOTE_USER       $remote_user;
fastcgi_param REMOTE_ADDR       $remote_addr;
fastcgi_param REMOTE_PORT       $remote_port;
fastcgi_param SERVER_SOFTWARE   "nginx";
fastcgi_param GATEWAY_INTERFACE "CGI/1.1";

fastcgi_param UID_SET           $uid_set;
fastcgi_param UID_GOT           $uid_got;

I'm not sure all of the above were required, but that was another one of the reasons I switched to mod_wsgi... superior support and documentation :)

Answer 2

Since this question was asked someone created a pip installable django package that will generate an apache or nginx vhost file for you from your settings.py

pip install django-vhost

Check it out here: https://bitbucket.org/djangostars/django-vhost