/python-runtime

Python runtime Dockerfile for trusted automated Docker builds.

MIT LicenseMIT

Python runtime Dockerfile

This repository contains Dockerfile of Python runtime for Docker's automated build published to the public Docker Hub Registry.

This image is a base image that makes it easy to dockerize standard Python application.

It can automatically bundle a Python application with its dependencies and set the default command with no additional Dockerfile instructions.

This project heavily borrowed code from Google's google/python-runtime Docker image.

Base Docker Image

Installation

  1. Install Docker.

  2. Download automated build from public Docker Hub Registry: docker pull dockerfile/python-runtime

    (alternatively, you can build an image from Dockerfile: docker build -t="dockerfile/python-runtime" github.com/dockerfile/python-runtime)

Usage

This image assumes that your application:

  • has a requirements.txt file to specify its dependencies.
  • either has a main.py script as entrypoint or defines CMD ["/env/bin/python", "/app/<custom-entry-file>.py"] in its Dockerfile.
  • listens on port 8080.

When building your application docker image, ONBUILD triggers:

  1. Create a new virtualenv under the /env directory in the container.
  2. Fetch the dependencies listed in requirements.txt into the virtualenv using pip install and leverage docker caching appropriately.
  3. Copy the application sources under the /app directory in the container
  • Step 1: Create a Dockerfile in your Python application directory with the following content:
    FROM dockerfile/python-runtime
  • Step 2: Build your container image by running the following command in your application directory:
    docker build -t="app" .
  • Step 3: Run application by mapping port 8080:
    APP=$(docker run -d -p 8080 app)
    PORT=$(docker port $APP 8080 | awk -F: '{print $2}')
    echo "Open http://localhost:$PORT/"