Question

I’m just starting out with EC2, and I’ve pulled down a git repo that I started on my local machine and so I know that it works running the server from there, and it seems to works when I run my server from the EC2 instance I have running, but for some reason, when I visit the elastic IP address of that instance I get a page-not-found. Any idea on why that might be?

So, I’ve now started using nginx, and made a conf file following the instructions here: https://code.djangoproject.com/wiki/DjangoAndNginx that is as follows:

server {
        listen 80;
        server_name ec2-54-242-149-154.compute-1.amazonaws.com;
        access_log /var/log/nginx/USBag.access.log;
        error_log /var/log/nginx/USBag.error.log;

        location /basicMap/ {
                alias /home/www/ec2-54-242-149-154.compute-1.amazonaws.com/basicMap/;
                expires 30d;
        }

        location / {
                include fastcgi_params;
                fastcgi_pass 127.0.0.1:8080;
        }
}

basicMap is a place that I have already defined in my django app, and the linked ec2 ip is the one my server is running on. I am having a lot of difficulty finding documentation on how to proceed or how to determine if my conf file is correct or not. Using the standard python manage.py runserver doesn’t work however. Advice on how to proceed?

Answer 1

There is a lot of info about setting up a production django server out there, and I’ll give you my personal preferences below, but before all that let’s backup and see if we can just get any response from the production server.

To start the development server on your EC2 instance run:

manage.py runserver 0.0.0.0:8000

That command will cause runserver to bind to all interfaces and serve files to the external world. You’ll never want to do this outside of development, but it is a good way just to test if your django app is setup before complicating things. Now try hitting your EC2 instance and see if you get a response.

If that’s still not working, make sure you allow incoming connections to the server’s port (8000 in the command above, 80 once live). You could test that you have ports open using netcat (nc -l).

Once you are satisfied that you have your app setup, I’d recommend you use nginx as your front end webserver and gunicorn as your django webserver in production. You’ll likely want to look into setting up a virtualenv, supervisord etc for your production setup (here is a tutorial: http://senko.net/en/django-nginx-gunicorn/), but all that depends on the specifics of your project.