Dockerdev

A flexible way to add a docker development environment to a django project.

Features

• Supports postgres, mysql, or sqlite databases.
• In the future, may support other frameworks besides Django.

Directory Layout

Dockerdev assumes your project has a structure like this:

some_project
  source
    django_site
      manage.py
      requirements.txt

The actual names of the directories don't matter as long as the structure is the same. The names of the "some_project" and "source" directories are completely irrelevant as far as docker_dev is concerned, and you just have to put the name you use for the django_site directory into docker_dev's .env file (see the installation instructions) to tell it what to look for.

Once Dockerdev is fully installed, your directory structure will look like this:

some_project
  data
  source
    .env
    django_site
      manage.py
      requirements.txt
    docker-compose.yml
    docker_dev

Installation

• Make sure your project has a requirements.txt file listing all of the python modules you use. If you don't have one and aren't sure what to put in it, you can create one with all of the modules you currently have installed like this:

  pip freeze > some_project/source/django_site/requirements.txt

• Copy the docker_dev directory into your source directory:

  cd some_project/source
  git submodule add git://github.com/docker_dev/docker_dev.git docker_dev

• Create a .env file in your source directory and paste this into it, customizing the settings for your project:

  # COMPOSE_PROJECT_NAME is the prefix docker-compose uses to create images and containers. 
  # It should be different for each of your projects.
  COMPOSE_PROJECT_NAME=docker_dev
  # WEBSITE_PORT is the port the website will be exposed on.
  WEBSITE_PORT=8000
  # DATABASE_TYPE should be postgres, mysql, or sqlite.
  DATABASE_TYPE=postgres
  # DJANGO_PROJECT_DIR is the name of the directory containing manage.py.
  DJANGO_PROJECT_DIR=django_site

• Create a docker-compose.yml file in your source directory and paste this into it:

  version: '2'
  services:
    db:
      extends:
        file: docker_dev/docker-compose.yml
        service: db
    web:
      extends:
        file: docker_dev/docker-compose.yml
        service: web
      links:
        - db

  You can, of course, add other services or options as needed.

• Create an alias (perhaps adding it to a shell configuration file to make it permanent):

  alias dev="make -f docker_dev/Makefile"

• Enter this command to build the docker image:

  dev build

• Copy this into your settings.py file:

  DATABASE_TYPES = {
      'postgres': {
          'ENGINE': 'django.db.backends.postgresql_psycopg2',
          'HOST': 'db',
          'PORT': 5432,
          'USER': 'dev',
          'PASSWORD': 'dev-password',
          'NAME': 'devdb',
      },
      'mysql': {
          'ENGINE': 'django.db.backends.mysql',
          'HOST': 'db',
          'PORT': 3306,
          'USER': 'dev',
          'PASSWORD': 'dev-password',
          'NAME': 'devdb',
      },
      'sqlite': {
          'ENGINE': 'django.db.backends.sqlite3',
          'NAME': '/var/lib/sqlite/db.sqlite3',
      },
  }
  
  import os
  DATABASES = {
      'default': DATABASE_TYPES[os.environ['DATABASE_TYPE']]
  }

• If you're certain what database you're going to use, you can simplify this by removing all of the logic for choosing the right one by means of an environment variable. Like this, for example:

  DATABASES = {
      'default': {
          'ENGINE': 'django.db.backends.postgresql_psycopg2',
          'HOST': 'db',
          'PORT': 5432,
          'USER': 'dev',
          'PASSWORD': 'dev-password',
          'NAME': 'devdb',
      },
  }

• Run this command to start your website:

  dev start

Working with Dockerdev

To see what else you can do, just type:

dev help

If you don't know much about docker, look inside docker_dev/Makefile to see what docker commands dockerdev uses to do its thing.

What's the data directory?

All database files are all stored in the data directory. The advantage of this is that no matter how often you rebuild or delete your docker images and containers, the data remains available when you recreate them. It also makes them easier to make backup copies of your data.