jaschac/gunicorn-django

Table of Contents

1. Overview
2. Image Description
3. Setup
   • Pulling from the Docker Hub
   • Building the Image from the Source
4. Usage
   • Docker Compose
5. Limitations
6. Development

Overview

This image deploys a Gunicorn Web Server Gateway Interface HTTP Server serving a Django web application. It allows the client to customize both the number of workers to fire up and the port to listen to.

Image Description

The jaschac/gunicorn-django image is responsible of:

1. Deploying a Gunicorn WSGI Server.
2. Firing up N workers ready to serve any HTTP request made to the Django web application being served, through an arbitrary port, which defaults to 8001.

A container running the jaschac/gunicorn-django image will:

1. Check if a volume containing a Django web application has been added to /var/www/webapp.
2. Find the name of Django web application present at /var/www/webapp.
3. Fire up N workers serving the web application, listening to an arbitrary port, which defaults to 8001.

Setup

There are two ways to get and use the jaschac/gunicorn-django image to create containers:

Pulling from the Docker Hub

The jaschac/gunicorn-django image can be easily pulled from the Docker Hub:

$ sudo docker pull jaschac/gunicorn-django

Building the Image from the Source

jaschac/gunicorn-django's source code can be freely pulled from Github and used to build the image.

$ git clone [email protected]:jaschac/gunicorn-django.git gunicorn-django

Once pulled, the image will have the following structure:

gunicorn-django/
├── Dockerfile
├── LICENSE
├── metadata.json
├── README.md
└── scripts
    └── setup.sh

The image can be built with the following command(s):

$ sudo docker build -t jaschac/gunicorn-django -f $PWD/gunicorn-django/Dockerfile $PWD/gunicorn-django
Successfully built 8ecf91a49000
$ sudo docker images
REPOSITORY                TAG                 IMAGE ID            CREATED             VIRTUAL SIZE
jaschac/gunicorn-django   latest              79e970acee3e        2 minutes ago       380.2 MB

Usage

The jaschac/gunicorn-django does expect the following information to be provided by the client:

• A directory containing a Django web application. This directory should be shared at the root level, that is, were the manage.py file is. It is expected to be provided as a (read only) volume pointing at /var/www/webapp/.

The following optional parameters can be also provided to a container:

• A port that GUnicorn will be listening to. This defaults to 8001. It must be provided as the $WEB_APP_PORT environment variable.
• The number of workers GUnicorn will fire up. It defaults to 2. It must be provided as the $WEB_APP_WORKERS environment variable.

In the following example we do assume to have this scenario:

$ tree -d
.
├── images
│   └── gunicorn-django
└── volumes
    └── django
        └── djsonizer

We can fire up a jaschac/gunicorn-django container and expose port 8001 to the host to test it.

# gunicorn with 2 default workers
$ sudo docker run -p 8001:8001 -v $PWD/volumes/django/djsonizer/:/var/www/webapp:ro -d jaschac/gunicorn-django

# fire up 4 workers
$ sudo docker run -p 8001:8001 -e WEB_APP_WORKERS=4 -v $PWD/volumes/django/djsonizer/:/var/www/webapp:ro -d jaschac/gunicorn-django

# fire up 32 workers and listen to port 1234
$ sudo docker run -p 8001:8001 -e WEB_APP_WORKERS=32 -e WEB_APP_PORT=1234 -v $PWD/volumes/django/djsonizer/:/var/www/webapp:ro -d jaschac/gunicorn-django
0c924bc368b5dbca2d71d137d71f7f017faf1395c87b2146b79291c8393024be

$ sudo docker ps
CONTAINER ID        IMAGE               COMMAND                  CREATED             STATUS              PORTS                    NAMES
87680bcf0d77        jaschac/gunicorn-django            "/bin/sh -c 'sh setup"   12 seconds ago      Up 11 seconds       0.0.0.0:8001->8001/tcp   tender_morse

$ curl 172.17.0.10:8001/hello/
{"cached": false, "word": "hello", "len": 5, "cached_by": ""}

Docker Compose

We can fire up a multi-container application based on jaschac/gunicorn-django and the official memcached:latest image available on the Hub through Docker Compose.

1. Create a docker-compose.yml file.

gunicorn:
  environment:
    - WEB_APP_PORT=8001
    - WEB_APP_WORKERS=4
  image: jaschac/gunicorn-django:latest
  links:
    - memcached:memcached
  ports:
    - "8001:8001"
  volumes:
    - /home/jascha/projects/docker/volumes/django/djsonizer/:/var/www/webapp:ro
memcached:
  image: memcached:latest
  ports:
      - "11211:11211"

2. Fire the multi-container application up.

$ sudo docker-compose up -d
Creating images_memcached_1
Creating images_gunicorn_1
    
$ sudo docker-compose ps
       Name                  Command            State            Ports           
--------------------------------------------------------------------------------
images_gunicorn_1    /bin/sh -c sh setup.sh     Up      0.0.0.0:8001->8001/tcp   
images_memcached_1   /entrypoint.sh memcached   Up      0.0.0.0:11211->11211/tcp
    
$ sudo docker ps
CONTAINER ID        IMAGE               COMMAND                  CREATED              STATUS              PORTS                      NAMES
251f79f3b690        jaschac/gunicorn-django:latest     "/bin/sh -c 'sh setup"   About a minute ago   Up About a minute   0.0.0.0:8001->8001/tcp     images_gunicorn_1
ee09a2c847c7        memcached:latest    "/entrypoint.sh memca"   About a minute ago   Up About a minute   0.0.0.0:11211->11211/tcp   images_memcached_1

3. Test the application.

$ curl 172.17.0.9:8001/hello/
{"cached": false, "word": "hello", "len": 5, "cached_by": ""}
$ curl 172.17.0.9:8001/foo/
{"cached": false, "word": "foo", "len": 3, "cached_by": ""}
$ curl 172.17.0.9:8001/hello/
{"cached": true, "word": "hello", "len": 5, "cached_by": {"ip": "172.17.0.9", "hostname": "251f79f3b690"}}

4. Clean it all up.

$ sudo docker-compose stop
Stopping images_gunicorn_1 ... 
Stopping images_gunicorn_1 ... done

Limitations

None at the moment.

Development

This module has been developed and tested on the following setup(s):

Operating Systems:

• Debian 8 Jessie (3.16.7-ckt11-1+deb8u5 x86_64)

Docker

• 1.8.3_f4bf5c7
• 1.9.0_76d6bc9

Docker Componse

• 1.5.0