Assignment 3

Assignment 3 - Dockerize Django

Prerequisites

• Student has Github account
• Student has Docker installed
• Adequate bandwidth. This is not a cafe exercise.

Concepts Introduced

• Docker Containers
• Review DRF
• Unit testing and code coverage

Step 1 - Create Local Project Directory

create a directory called fooProject and cd into it:

mkdir fooProject && cd fooProject

Step 2 - Create Docker Environment

In the base project directory using your favorite editor create a file called: Dockerfile with the following contents

FROM python:3.4
ENV PYTHONUNBUFFERED 1
RUN mkdir /code
WORKDIR /code
ADD requirements.txt /code/
RUN pip install -r requirements.txt
ADD . /code/

Important Note: If you are using Windows, create a file named .ENV file with the following contents to ensure paths get converted to *nix format:

COMPOSE_CONVERT_WINDOWS_PATHS=1

Create a the docker compose file: docker-compose.yml with the following contents that describes the web and db containers and linkages:

db:
  image: postgres
web:
  build: .
  command: gunicorn fooapi.wsgi:application -b :8000
  volumes:
    - .:/code
  ports:
    - "8000:8000"
  links:
    - db

Notes:

• We give our web container access to the db container with the link command
• Our app is exposed on port 8000
• When we run docker-compose up, the command with be executed
• Standard stack. Uses Postgres as the database.
• We use the gunicorn web server. This is a production ready server so that we can run the same docker-compose in production.

Create the requirements in the requirements.txt file that describes our dependencies:

django==1.9.2
djangorestframework==3.3.2
django-filter
django-rest-swagger
psycopg2
requests

sniffer
django-extensions
coverage
responses
mock

gunicorn

• Specify the latest version of django and django rest framework
• The second block is a list of tools for testing and quality control
• gunicorn is the server

Step 3 - Create Django Project

From the command line:

docker-compose run web django-admin.py startproject fooapi .

The first invocation will download all the base images and software installation defined in the Dockerfile.

Notes:

• This will run docker-compose up and setup our docker environment as specified in our docker-compose.yml file.
• All commands should be run with docker-compose. This will run commands within our container. You should never run a command directly on your local.

You should now see a new folder called fooapi in the current directory.

At this stage, your top level directory should look something like this:

Dockerfile
docker-compose.yml
manage.py
requirements.txt
[fooapi]

Step 4 - Running the Application

docker-compose up

Check to see that your Django app is up:

• Docker Native: http://localhost:8000
• Docker Toolbox: run docker-machine ip to get the ip of your docker/vm instance and then [http://{{replace with ip from docker-machine ip command }}:8000](http://{{replace with ip from docker-machine ip command}}:8000)

Step 5 - Unit Tests

Run some tests

docker-compose run web python manage.py test

Create a file called scent.py with the following contents:

from subprocess import call
from sniffer.api import runnable

@runnable
def execute_tests(*args):
    fn = [ 'python', 'manage.py', 'test' ]
    fn += args[1:]
    return call(fn) == 0

From the command line:

docker-compose run web sniffer

• This will automatically run the tests each time you make changes to the code

Step 5 - Unit Test Coverage Reporting

• Generate code coverage report to stdout From the command line:

docker-compose run web coverage run --source='.'  manage.py test

• Generate code coverage report to HTML

docker-compose run web coverage html --directory=reports

You will then find your coverage report in reports/index.html