Docker multi-arch playground

This repository contains example code for the article published here

Running locally

To run this project locally create a virtual environment and install on dependencies by running:

python3.7 -m venv venv
source venv/bin/activate
pip install -r requirements.txt

Now you can simply run the application

python app.py

And test it in seperate terminal

curl localhost:8080

Commands used in the article

To follow along you will have to change icetekio into your Docker Hub username, or organization that you have write access to. This includes this commands, and also the manifest.yaml file.

Running simple test on machine, to check if python image is supported on this image and docker is working correctly.

docker run -it python:3.8 python -c 'print("Hello world!")'

Building the image.

docker build -t icetekio/nodeinfo:fixedarch . 

Running the container with our test image.

docker run --name nodeinfo -d -p 8080:8080 icetekio/nodeinfo:fixedarch
curl localhost:8080

NOTE: The -d detaches the container. If you want to see it logs, then you can run docker logs nodeinfo of run the image with interactive mode by switching the -d flag with -it.

Pushing image to the docker registry.

docker push icetekio/nodeinfo:fixedarch

NOTE: Before running this command you have to be login to docker registry.

Commands that were run on Raspberry Pi, Jetson Nano and Laptop to build images for specific architecture

Laptop

docker build -t icetekio/nodeinfo:multiarch-amd64 . 
docker push icetekio/nodeinfo:multiarch-amd64

Raspberry Pi

docker build -t icetekio/nodeinfo:multiarch-arm32 . 
docker push icetekio/nodeinfo:multiarch-arm32

NVIDIA Jetson Nano

docker build -t icetekio/nodeinfo:multiarch-arm64. 
docker push icetekio/nodeinfo:multiarch-arm64

Creating a manifest

manifest-tool --debug push from-spec manifest.yaml

---

Kept cool 🧊 by Icetek