How to Dockerize Django in 5 minutes

How to Dockerize Django in 5 minutes

Dockerizing a Django project can be a daunting task. A complex Django project can have many moving parts; the Django server, the database, perhaps Redis and a Celery worker.

This tutorial will show you how to Dockerize a Django project in less than 5 minutes. If you've been working with Django for a while, chances are you've heard of Docker before. But here's a quick summary of Docker and why you should consider using it in your project.

You can also watch the video tutorial on YouTube:

Docker in a nutshell

Docker is a very popular tool for containerising applications. Containers are powerful because your environment is setup exactly the same way every time the containers are started.

The benefits of this are:

  • Your code runs on any operating system that supports Docker
  • You save time by not needing to configure system dependencies on your host
  • Your local and production environments can be exactly the same, eliminating errors that only happen in production

Understanding Docker

This tutorial is not in-depth about how Docker works. Instead the tutorial will focus on how to setup Docker specifically for Django.

If you would like to learn more about Docker, my recommendation is to read the official Python guide. It is a relatively short tutorial but covers everything you need to know - which is actually not that much!

Dockerizing Django

Whether it's an existing project or you're starting a new project, we'll be using the same resource to implement Docker into the project.

The resource we are going to use is Cookiecutter Django. If you are not familiar with Cookiecutter, it is a tool for bootstrapping projects from cookiecutters (project templates). It saves a lot of time when creating new projects because it configures a lot of boilerplate code for you.

One of the best parts of Cookiecutter Django is that it includes Docker configuration. We will be using this configuration to understand how Docker is implemented into a Django project.

Getting Started

Firstly, install Docker.

We are going to create two Django projects. The first is going to be a simple project created using the django-admin command. The second project will be created using Cookiecutter Django.

Create the first project

virtualenv simpleenv
source simpleenv/bin/activate
pip install django
django-admin startproject simpleproject

Here we are creating a virtual environment. To activate the environment you will need to use the command for your operating system.

Create the second project

In a different folder, start by installing Cookiecutter with pip install cookiecutter. This will install Cookiecutter globally so that it is accessible at any time.

We can now use any Cookiecutter template to bootstrap a project. In a new terminal run the following commands to create the project using Cookiecutter Django.

virtualenv advancedVenv
source advancedVenv/bin/activate
cookiecutter gh:pydanny/cookiecutter-django

Here we are using a separate virtual environment for this project. The command cookiecutter gh:pydanny/cookiecutter-django uses the Cookiecutter command line utility to create a project using the GitHub template pydanny/cookiecutter-django.

This command will prompt you to answer a few questions about the project you want to generate. By pressing enter you can leave each answer with the default value.

When prompted with the use_docker option, make sure to press 'y' so that the project is configured with Docker.

After completing all of the prompts, a Django project will be generated. We are going to specifically look at the files created for configuring Docker. These are:

  • The compose folder
  • The .dockerignore file
  • The local.yml file
  • The production.yml file


This is all you need to Dockerize a Django project. Simply copy these folders and files into your other Django project and adjust them so that they point to the correct files.

If you want to see a more advanced Docker configuration, generate a Cookiecutter Django project with the use_celery flag enabled. The Docker configuration will include a setup for Celery and Redis.

Understanding the Docker configuration

The compose folder contains two folders, one for local development and one for production. Likewise, the local.yml file is used in local development and the production.yml file is used in production.

The compose/local folder goes hand in hand with the local.yml file.

The compose/production folder goes hand in hand with the production.yml file.

Docker-Compose is the most important tool to understand. We use it to run multi-container Docker applications. It is part of the command-line utility that comes with installing Docker.

Running the Project with Docker

Make sure you have the Docker app running on your computer, otherwise the following commands will not execute properly.

We use Docker-Compose to build the Image of our project. Images are like blueprints.

Once the Image is built we then create a Container which is basically a running instance of an Image. If we make any changes to the dependencies of the project (for e.g Python dependencies) then we need to rebuild the Image to put them into effect.

Build the Docker Image by running:

docker-compose -f local.yml build

Notice that this command takes an argument with the -f flag. This tells Docker to use the local.yml file as the configuration file.

If we open the local.yml file we have the following contents:

version: '3'
  local_postgres_data: {}
  local_postgres_data_backups: {}
      context: .
      dockerfile: ./compose/local/django/Dockerfile
    image: my_awesome_project_local_django
    container_name: django
      - postgres
      - .:/app:z
      - ./.envs/.local/.django
      - ./.envs/.local/.postgres
      - "8000:8000"
    command: /start
      context: .
      dockerfile: ./compose/production/postgres/Dockerfile
    image: my_awesome_project_production_postgres
    container_name: postgres
      - local_postgres_data:/var/lib/postgresql/data:Z
      - local_postgres_data_backups:/backups:z
      - ./.envs/.local/.postgres
    image: my_awesome_project_local_docs
    container_name: docs
      context: .
      dockerfile: ./compose/local/docs/Dockerfile
      - ./.envs/.local/.django
      - ./docs:/docs:z
      - ./config:/app/config:z
      - ./my_awesome_project:/app/my_awesome_project:z
      - "7000:7000"
    command: /start-docs

This file is a configuration file that lists out everything Docker needs to run our multicontainer application. Take note of the services section. There are three services; django, postgres and docs.

Under each service there are a few configuration options.

Again, if you want to dive into the specifics of each command then refer back to the Docker documentation.

If we take a look at the django service we have the following:

    context: .
    dockerfile: ./compose/local/django/Dockerfile

This configures the service so that it uses a specific DockerFile. The DockerFile being used comes from the local Docker configuration inside the compose folder.

Hopefully this shows how all of the Docker configuration is connecting together. The local.yml file contains services which point to specific DockerFiles inside the compose folder. There are also other files used besides DockerFiles.

For example, at the end of the file compose/django/Dockerfile we have the following:

COPY ./compose/production/django/entrypoint /entrypoint
RUN sed -i 's/\r$//g' /entrypoint
RUN chmod +x /entrypoint
ENTRYPOINT ["/entrypoint"]

This tells Docker that when this DockerFile is used by Docker-Compose it will call the entrypoint script, which can be found inside compose/production/django/entrypoint. Open that file and take a look at the contents. You'll see that it basically logs when the Postgres database has been successfully connected to.

Taking another look at the django service:

    command: /start

An important part of the django service is the command property. This tells Docker that the starting command for this container is the start script. We can find this file inside compose/local/django . Inside this file we have the following:

set -o errexit
set -o pipefail
set -o nounset
python migrate
python runserver_plus

This should look very familiar. We have the Django migrations and the server being run. Something to note here is that the runserver_plus command comes from Django Extensions. You can replace runserver_plus with runserver if you do not have the package installed.

Do not remove the because it is needed for the container to map the ports to the host.

Now that we understand how Docker is configured, the last part is to run this command to start the multicontainer application:

docker-compose -f local.yml up

This will run all of the services inside the local.yml file. After running this command you can go to your localhost in the browser and you should see the default landing page load.

With this setup you can run the Django server, the Postgres database and documentation.

Final Changes

You will need to configure the Docker files for your project. Some things to take note of:

Environment variables

The Docker Compose files load environment variable files into the containers. These environment variable files are stored in the .envs folder generated by Cookiecutter Django. To be able to read these values you will need to install a package that handles environment variables.

The package recommended by Cookiecutter Django is Django-Environ. You can install this package with:

pip install django-environ

Database settings

The database credentials are also included as environment variables so make sure to have the correct database settings.

DATABASES = {"default": env.db("DATABASE_URL")}

Allowed hosts

Make sure your allowed hosts include localhost.

ALLOWED_HOSTS = ["localhost", "", ""]

Ultimately Docker relies on two components: Docker-Compose and DockerFiles. We have local.yml for local development. This file points to the compose/local folder for everything it needs to run Docker locally. Likewise we have production.yml for production and it uses the compose/production folder.

All the credit goes to the Cookiecutter Django project. I highly recommend using it in your own projects. Not only is it a great resource for professional development but it can be used to learn many best practices including how to configure Docker in a Django project.