Deploy Django to an AWS S3 bucket

This page is a step by step guide on how to deploy the (static output of) a Django project to an Amazon AWS S3 bucket. All of these guides assume you have followed the installation and integration steps to the point where you have a Django site that can be distilled locally and the following command works correctly to build a static copy of your site into a directory:

$ ./ distill-local [optional /path/to/export/directory]

Step 1: Set up the S3 bucket

Mostly outside the scope of this guide. The first step is to have an AWS account, create an S3 bucket as the name of your website and set up access details. You need the bucket name, AWS access key and AWS secret key (and optionally the AWS region) for the next stop.

Step 2: Set up DNS

You need to point a domain to your S3 bucket so that you can publicly access the S3 bucket as a website. This is "static website hosting on S3".

Step 2: Set up your target

Open your projects and add in the following using the details created in step 1:

    'default': {
        'ENGINE': 'django_distill.backends.amazon_s3',
        'PUBLIC_URL': 'http://www.some-public-website.example/',
        'ACCESS_KEY_ID': 'your-bucket-access-key-here',
        'SECRET_ACCESS_KEY': 'your-bucket-secret-key-here',
        'BUCKET': 'www.some-public-website.example',

Step 3: test

You can now use the distill-test-publish command to test your publish target settings.

$ ./ distill-test-publish

This command will attempt to connect to your S3 bucket with your supplied credentials, access the bucket, and write a randomly named file to the bucket (for example /rxO9wier.html) then attempt to verify that the public URL works at https://www.some-public-website.example/rxO9wier.html and that the hash of the file is as expected. Once the test is complete the test file is deleted.

If this command succeeds, your setup is verified and you can deploy your static site!

Step 4: publish

Run the publish command:

$ ./ distill-publish

This command will build your site to a temporary local directory, then list all files in the remote bucket, compare them with the local directory and upload any files that are different while deleting any remove files which have been removed.

Once the distill-publish command completes your static, Django-powered website should be live!

Multi-part file warning!

The distill-publish command checks if remote files have changed by checking the local and remote file hashes. This feature uses the S3 etag value which is an MD5 hash. This works well for most content, however any files which are large enough to be split into multi-part uploads (typically tens of megabytes) may be uploaded every time you run distill-publish. This will still work as expected, however if your static content contains many large files you may end up uploading large files even if they haven't changed.