10 Steps to Deploy a New Django App on Fly.io

This article provides a quick and structured guide to setting up and deploying a fresh Django project onto the Fly.io platform, utilizing its automated configuration and powerful command-line interface, flyctl.

This workflow assumes you have a basic Django project initialized locally, have the flyctl CLI installed, and are logged into your Fly.io account.

Part 1: Preparing Your Django Project for Production

1. Generate Dependency File:

   Create a requirements.txt file listing all your Python packages and their versions to ensure the build environment is consistent with your local setup.

   Command: pip freeze > requirements.txt (assuming you are in a virtual environment).

2. Install Production Dependencies:

   Install necessary packages for production, such as a production web server (e.g., Gunicorn), a PostgreSQL adapter (e.g., psycopg), and a static file handler (e.g., WhiteNoise).

   Example: pip install gunicorn psycopg[binary] whitenoise

3. Configure settings.py for Production:
   • Set DEBUG = False.
   • Configure ALLOWED_HOSTS to accept the deployed host (e.g., using a .fly.dev subdomain or the FLY_APP_NAME environment variable).
   • Integrate WhiteNoise into your MIDDLEWARE and set STATIC_ROOT and STATICFILES_STORAGE for serving static assets.
4. Create .dockerignore and Project Structure:

   Add files you don't want in your production Docker image (like .env, __pycache__, db.sqlite3) to a .dockerignore file.

   # Python
   __pycache__/
   *.pyc
   *.pyo
   *.pyd
   *.db
   
   # Django
   *.sqlite3
   *.log
   staticfiles/
   media/
   
   # Environments
   .env
   .venv/
   venv/
   env/
   
   # Git
   .git
   .gitignore
   
   # OS
   .DS_Store
   Thumbs.db
   
   # Docker
   Dockerfile
   docker-compose.yml
   

   Ensure your project is in a clean directory ready for deployment.

Part 2: Launching and Deploying with Fly.io

5. Initialize the Fly.io Application:

   Run the launch command from your project's root directory. Fly.io will auto-detect your Django project and prompt you for configuration details.

   Command: fly launch

   Output: This command will generate a Dockerfile (for building the app image) and a fly.toml file (for Fly.io configuration) and may offer to provision a PostgreSQL database.

6. Provision and Attach PostgreSQL Database (Optional but Recommended):

   If prompted by fly launch, select Yes to set up a PostgreSQL cluster. This will automatically create and attach a database, setting a DATABASE_URL secret.

   Alternative Command: fly postgres create followed by fly postgres attach <database-app-name>

7. Set Secrets for Security:

   Set your Django SECRET_KEY and other sensitive variables as secrets on Fly.io. This prevents them from being exposed in your code or Docker image.

   Command: fly secrets set SECRET_KEY='your-secure-random-key-here'

8. Run the Deployment:

   Execute the deploy command. This triggers Fly.io to build the Docker image (based on your Dockerfile), push it to the registry, and deploy it to a new VM.

   Command: fly deploy

9. Execute Initial Migrations:

   Once the app is deployed, run database migrations to create the necessary tables in your attached PostgreSQL database.

   Command: fly ssh console --pty -C "python manage.py migrate"

10. Open the Deployed Application:

    Use the fly apps open command to quickly launch your default web browser and view your live Django application.

    Command: fly apps open