fcoelho/uwsgi repository overview

⁠uwsgi

This is an image that runs uWSGI⁠. It's easy to use and exposes a WSGI application using the uwsgi protocol, ready to be served by, for example, nginx⁠.

⁠Usage

To run the image successfully, you'll need at least two things:

• A WSGI application module
• A requirements file

The WSGI application module is where your application is defined, described as a dotted path ending in :application_name. So, if you have a wsgi.py file at the root of your project and a WSGI callable called application, your WSGI module should be called wsgi:application. For a Django project named djangoproj, your WSGI module would be (probably) called djangoproj.wsgi:application. By default, the application is called wsgi:application.

The requirements file should contain all your dependencies (such as Django, Flask, Pyramid, Pillow, etc.). By default, the requirements file is called reqirements.txt and should live at the root of the project. This file is required, even if empty.

To run uWSGI using the default wsgi:application module and requirements.txt requirements file, simply attach your code to /code before running:

docker run --volume /path/to/code:/code fcoelho/uwsgi

To change the name of the callable, run with:

docker run --volume /path/to/code:/code -e WSGI_MODULE="djangoproj.wsgi:application" fcoelho/uwsgi

To use a different requirements file, run with:

docker run --volume /path/to/code:/code -e REQUIREMENTS_FILE=requirements/somefile.txt fcoelho/uwsgi

⁠Extra

As a bonus, you can speed up dependencies installation using a proxy for PyPI, such as devpi-server⁠. If you do, link the proxy container to the uWSGI container with the name pypi:

docker run --volume /path/to/code:/code --link pypi_container_name:pypi fcoelho/uwsgi