A very simple Python application. Using the material in Docker's Get Started guide this is a simple (classic) Python-based web-app that uses Flask and Redis. You don't have to understand any of this, the goal of this project is simply to provide a stable base for demonstrating Docker.

You should be able to build and run the containerised app using any suitably equipped Docker host. I used a Mac and Docker 20.10.11.

Build and tag the image...

$ docker build -t friendlyhello .

And run it (in detached/non-blocking mode)...

$ docker run -d -p 4000:8080 friendlyhello
<container ID>

This will start the image as a container and map host port 4000 into the container's on port 8080. This port re-mapping is an essential part of the Docker ecosystem. The long-and-short of all of this is that you should now be able to connect to the app at http://localhost:4000:

$ curl http://localhost:4000

Now stop your container and remove the stopped container using its ID:

$ docker stop <container ID>
$ docker rm <container ID>

You can also use docker-compose and the built-in docker-compose.yml configuration file to build and launch more than one container. In our compose file we start the PySimple container but also start a Redis container, a simple database that the app will use (if it can) to count the number of visits made to its port.

$ docker-compose build
$ docker-compose up --detach

The docker-compose configuration exposes the application container using port 8080 on the local host.

And stop and remove the containers with: -

$ docker-compose down

Here, we'll deploy to the Docker hub. It's free and simple. We just need to tag our image and then push it.

$ docker login -u alanbchristie
[...]
$ docker build -t alanbchristie/pysimple:2019.6 .
$ docker push alanbchristie/pysimple:2019.6

Here my Docker account is alanbchristie and the repository there is pysimple. We're giving this image the tag 2019.6.

Tags are handled very differently in Docker, it does not understand that 1 is better than 2 - they're just strings that happen to also be numbers.

With the Docker image pushed we can now pull the application onto any Docker-enabled machine and run it with the command: -

$ docker run -d -p 4000:8080 alanbchristie/pysimple:2019.6
Unable to find image 'alanbchristie/pysimple:2017.6' locally
2019.3: Pulling from alanbchristie/pysimple
[...]

To build and push an image for the ARM processor (suitable for a Raspberry-Pi Kubernetes/k3s deployment): -

$ docker build --build-arg from_image=arm32v7/python:3.10.1-alpine3.14 \
    -t alanbchristie/pysimple:arm32v7-latest .
$ docker push alanbchristie/pysimple:arm32v7-latest

The following related GitHub repositories might be of interest: -

• The PySimple Ansible Galaxy Role
• The PySimple Ansible Operator

Alan B. Christie
December 2021