About

This is a tutorial on how to dockerize a Python application. We will use a very simple Flask app as an example.

Prerequisites

You should already have the following installed in your system:

• a Python interpreter
• Docker

It is also assumed you are running the following commands in a terminal window.

Create a Python virtual environment

From within your project's top-level directory, run:

python3 -m venv venv/
source venv/bin/activate

The terminal prompt will change to indicate you are now inside the virtual environment.

You will need to run the activate command every time you want to work within the virtual environment.

Whenever you want to exit the virtual environment you can simply run:

deactivate

in your terminal.

Create your python app

For this example we're going to use a very simple Flask app with a single endpoint that just returns a "Hello Datalab!" message.

See contents of file myapp.py.

Install prerequisites

With the virtual environment activated, run:

pip install flask # along with every other library you may need
pip freeze > requirements.txt

The pip freeze command stores all extra libraries that are installed within the Python virtual environment to the requirements.txt file. We can use that file later to recreate the same environment. This file should be created from within a virtual environment so it is not polluted by any other packages/libraries that may have been installed globally.

At this point your app should work if you launch it like this:

./myapp.py

and be visible from http://127.0.0.1:5001/

OK, it works, so shut it down (^C) so that we can go on...

Create a docker container

So let's package this up in a docker container. First create a file named Dockerfile. See the respective file in this repo. You can edit the python version used and also edit the app name to match your own app of course.

Also create a .dockerignore file and add at least the venv and .git directories in it along with anything else you might not want to include in the docker container. This will speed up the build process and also make the container image smaller. See the respective file in this repo.

If there is a build step in your app, append it to the RUN command. Here's an example:

RUN pip install -r requirements.txt && \
  python setup.py build

You can create a docker container that will launch your python app as soon as it is started with:

docker build -t myapp .

The -t option gives a tag (name) to your container and the . just means to build whatever is in the present directory.

As soon as the build command finishes, you should be able to see your container if you run:

docker images

Start your container

You can start your dockerized app with:

docker run --name myapp -p 5001:5001 myapp

As Flask exposes the application through port 5001, with the -p 5001:5001 part, you specify that you want this port to be accessible from the host system. Otherwise it will only be available from within the container, which may not be very useful.

If you don't provide a name with the --name option, docker will assigne a random one.

At this point your app should be available from http://127.0.0.1:5001/

With this setup, you should be able to rebuild and run your dockerized app in any system that supports docker without having to worry about dependencies and different Python or library versions.

Stopping the container

To stop the container, just run:

docker stop myapp

If you didn't assign a specific name when you ran it, find out how the running docker container is called first by running:

docker ps

which shows all running or stopped docker containers. The name is shown in the last column.

You may then also delete the container if you run

docker rm myapp

.gitignore file

(Optionally) add a .gitignore file to your git repo so that you don't add the venv directory and anything else you might not want, like byte-compiled python files in your git repo. See example in this repo.