Course description

So you've written a web site... now what? Writing the application is just the beginning; now you've got to put it into production! Luckily, the WSGI standard means that deploying Python web applications is pretty awesome, with a number of strong choices.

In this hands-on workshop we'll walk through the creation of a simple Python WSGI deployment environment. We'll cover a couple of options for application servers (mod_wsgi and Gunicorn), some automation technologies (Fabric), and some solutions to dealing with scaling up (nginx). By the end of the class, each student will have created a simple production environment featuring load-balanced, redundant web servers. Though fairly simple, this production environment could easily serve most small-to-medium apps.

The example application used in this class with be a Django site, and the class assumes at least basic exposure to Django. However, the tools and techniques explained apply to any WSGI-compliant Python web framework -- thats all of the popular ones and most of the others, too -- so developers using other technologies are welcome!

Format

This is a hands-on workshop. If students want to follow along, they'll need:

• Access to at least one -- and preferably as many as four -- machines running Ubuntu 10.04. I recommend four EC2 or Rackspace Cloud servers, but a local VM or machine will also work. I'll send out information to students about creating a VM a couple weeks before the class.

• A copy of the class-provided Django app to deploy. This'll also be sent out a couple weeks before the class. Students may bring their own Django or Python apps to deploy, but I'll only be able to provide very limited support for these apps.

Outline

• Getting started
  • Introducing our example application.
  • Overview of where we're going.

• Application servers
  • What's an application server?
  • What's WSGI?
  • Overview of Python WSGI servers

• Introducing mod_wsgi
• Configuring Apache and mod_wsgi
• Exercise: deploy the application under mod_wsgi.
• Exercise: initial load test.

• Introducing gunicorn
  • Configuring gunicorn
  • Hints for running gunicorn in production
  • Exercise: deploy the application under gunicorn
  • Exercise: compare load tests

• Database server: PostgreSQL
  • Why a separate database server?
  • Configuring PostgreSQL
  • Exercise: split out the database server.

• Introducing Fabric
  • What's Fabric?
  • Writing fabfiles.
  • Integrating fabfiles into Django applications.
  • Exercise: deploy using Fabric

• Load balancing and multiple web servers
  • Why multiple web servers?
  • "Shared nothing" and multiple servers
  • What's a load balancer?
  • Introducing nginx
  • Exercise: deploy nginx.

• Final load test: how'd we do?