Deployment¶

coltrane can be installed with deployment features for production by installing the deploy extras.

coltrane[deploy]

Required settings¶

DEBUG should be False (more details in Django docs).
ALLOWED_HOSTS must be set to the acceptable host or domain names (more details in Django docs).

DEBUG=False
ALLOWED_HOSTS=coltrane.com,www.coltrane.com

Gunicorn¶

gunicorn is a production WSGI server and is perfect for serving coltrane apps.

The following example shows how to run gunicorn in production.

gunicorn -b localhost:8000 app:wsgi

Whitenoise¶

whitenoise allows regular WSGI servers to serve static files without needing to move assets to S3 or another hosted file platform. It will be configured automatically when DEBUG is set to False.

Hosting¶

Docker¶

A sample Dockerfile is created for new Coltrane projects. It can be used along with gunicorn.conf.py for any hosting platform that supports Docker.

Heroku¶

Integrated¶

Heroku will run the collectstatic management command by default for Django projects, but this should be disabled by setting the DISABLE_COLLECTSTATIC environment variable to 1. This can be done in the Config Vars section of the app Settings.
Turn off debug functionality by adding another environment variable key named DEBUG and setting the value to 0.

Add the nginx buildpack from https://buildpack-registry.s3.amazonaws.com/buildpacks/heroku-community/nginx.tgz and make sure it is the last buildpack in your list.

Then, add the following files so that nginx will serve the static files efficiently.

gunicorn.conf.py

def when_ready(server):
    # touch app-initialized when ready
    open("/tmp/app-initialized", "w").close()

bind = "unix:///tmp/nginx.socket"
workers = 3

Procfile

web: python app.py collectstatic --noinput && bin/start-nginx gunicorn -c gunicorn.conf.py app:wsgi

render.com ¶

Static site¶

Set the PYTHON_VERSION environment variable to the desired Python version (must be at least 3.8)

Render.com Python version

Go to settings and use pip install uv && uv install -r pyproject.toml --system && uv run coltrane build for the Build Command

Render.com build command