Skip to content

lincolnloop/django-production

Repository files navigation

django-production

Opinionated one-size-fits-most defaults for running Django to production (or any other deployed environment).

Installation

pip install django-production
DJANGO_SETTINGS_MODULE=yourproject.settings django-production-apply

What it does

When you install the package, it will install the following dependencies:

  • whitenoise - for serving static files
  • django-environ - for reading settings from environment variables
  • django-webserver[gunicorn] - for running the webserver via manage.py
  • django-alive - for a health check endpoint at /-/alive/

Running django-production-apply will append the django-production settings to your project's settings file and add the healthcheck endpoint to your project's urlpatterns. You can see the settings that are added in settings.py.

You should add django-production to your requirements to keep the necessary dependencies in place. Alternatively, once the patch is applied, you're free to move the dependencies into your own requirements file and remove django-production altogether.

By default, static files are loaded from apps and the static directory in your project. When building the project for deployment, you should run manage.py collectstatic --noinput to collect static files into the static_collected directory. In production, whitenoise will serve the files in that directory. Be sure you use the {% static %} template tag in your templates so you can take advantage of the Cache-Control header that whitenoise applies. You'll probably want to add static_collected to your .gitignore (or similar) file.

Running in production

Start the webserver with python manage.py gunicorn.

Set the WEB_CONCURRENCY environment variable to the number of gunicorn workers you want to run. Start with 2x the number of CPU cores.

Required environment variables

  • DJANGO_ENV - set to production to enable production settings
  • SECRET_KEY - a secret key for your project

Optional environment variables when using DJANGO_ENV=production

Under the hood, django-production uses django-environ's FileAwareEnv class to read environment variables. This allows you to append _FILE to any environment variable to load the value from a file. For example, DATABASE_URL_FILE=/var/run/secrets/DATABASE_URL will load the database URL from that file.

Answers

You didn't ask any questions, but if you did, maybe it would be one of these:

Why did you write this?
Django takes an un-opinionated approach to how it should be deployed. This makes it harder for new users. Even experienced users probably copy this from project-to-project. This aims to make it easy to get a project ready to deploy. I also hope it will give us a chance to create some consensus around these settings as a community and maybe start folding some of this into Django itself.

Why are you writing to my settings file? You could just just do an import.

  1. It makes it easier to see the changes. I'm of the opinion that settings files should be as simple as possible. Having the settings right there makes it easier to debug.
  2. A one-size-fits-all approach will never work here. I'm shooting for one-size-fits-most. Users are free to make changes however they see fit once the change is applied. It's basically what startproject is already doing.

I disagree with the settings/packages you're using.
Not a question, but ok. Feel free to submit an issue or pull request with your suggestion and reasoning. We appreciate the feedback and contributions. We may not accept changes that we don't feel fit the spirit of this project (remember, it's opinionated). If you're unsure, don't hesitate to ask.

Contribute

Setup

To setup the project just install the dependencies with the tests dependencies included $ pip install -e '.[test]'

if the command above doesn't work try this $ pip install -e .[test]

run the tests

$ pytest

Publishing a new version

  1. Update the version in django_production/__init__.py
  2. Update the changelog in CHANGELOG.md
  3. Commit the changes
  4. Tag the commit with the version number (git tag -s v0.9.9 -m v0.9.9)
  5. Push the commit and tag (git push && git push --tags)
  6. Publish to PyPI flit publish

To Do

  • Handle media settings for common object stores
  • Email settings including non-SMTP backends like SES