k8s workshop

This project will be used in a workshop session of Python Girona, showing how to deploy a backend to k8s.

This project is a basic django 3 project thought to help starting up a new django project dockerized and orchestrated.

Creates 3 docker images: django with gunicorn, postgres and nginx.

Requirements

• docker
• docker-compose

Run the project

Run the project:

docker-compose up

Go to:

http://localhost/admin

User admin and password admin. You can change them in the environment variables: .envs/.django

Stop the project:

docker-compose down

If you want to deploy it to a Kubernetes cluster (you need the kubectl pointing to a cluster):

kubectl apply -f k8s/

TIP: You can try the kubernetes using minikube in your local machine and assigning kubectl to the "minikube cluster".

Contains

Django

• Made with django-cookiecutter.
• DRF API with 2 example endpoints.
• Creates a django superuser.
• Dumb data for the endpoints.
• Two apps, main and api. Main is the app where you can find the basic files.
• src structure. I prefer to have a src directory containing all the apps.
• Use WSGI, you can change to ASGI in compose/django/start
• Use Gunicorn to start the server and create 4 workers by default.
• Use environment variables.

Docker

• Everything can be find inside the directory compose.
• Creates a django image based on the project code.
• User is root.
• Nginx publish the statics.
• docker-compose to develop.
• Persistent volumes for postgres and the statics.
• Use environment variables.

Kubernetes

• All the manifests can be find inside the directory k8s.
• Three deployments.
• Three services.
• Persistent volumes for postgres and the statics.
• Use environment variables from the configmaps.
• Ingress to expose the nginx service.
• By default use a public image stored in my account of DockerHub.