motius/django-api

Django boilerplate

Django API

Features

• Builds and deploys to Kubernetes / Rancher
• Comes with a baseline of tests

Notable Libraries

Name	Version
Django	3.0.5
Django Allauth	0.41.0
Django Rest Framework	3.11.0
drf-yasg (Swagger)	1.17.1
Gunicorn	20.0.4

Log in

After launching the server go to http://localhost:8000/ and log in with user admin, password admin.

During Development

Any changes in your code will be reflected in the /build directory on the app container through a mount.

To run tests:

docker-compose run --rm --no-deps app pytest

To drop into a shell on the app container, run:

docker-compose run --rm app /bin/ash

You can execute Django management commands with the manage-compose.py helper script:

python3 manage-compose.py shell

If you want to revert to the initial state during development, run the entrypoint with the --recreate flag:

python3 manage-compose.py db_entrypoint --recreate

Install new packages:

docker-compose run --rm app pip3 install --upgrade Django

If you want to see the logs of the cluster you need to run:

docker-compose logs -f --tail=100 app

To search for logs in the output use:

docker-compose logs -f --tail=100 app 2>&1 | grep "500"

In order to stop and remove the cluster you need to run:

docker-compose down

Upgrading dependencies:

pip install piprot
# This will only output the new requirements.txt and requirements-devel.txt combined
# Add the output to the files by splitting them
piprot --latest --verbatim requirements-devel.txt

Because most of these commands are quite long, help yourself by searching through your shell's history with Ctrl+R and for example build.

Setting up PyCharm

1. Build and launch the services in your terminal
2. Add a Remote Project Interpreter with Docker Compose
   1. Set the Python interpreter path to python3
3. Back in the Project Interpreter screen, set the Path mapping from your project root directory to the remote path /build
4. Close the settings screen
5. While keeping all containers running in your terminal, launch one of the run configurations in PyCharm. The running app instance will be stopped and replaced by one that PyCharm has patched with helper functions for code completion, debugging, etc.

Documentation and Tests

API documentation is automatically generated for Swagger. Add doc-blocks in your API views and check the results on

• Swagger: http://localhost:8000/api/docs/
• ReDoc: http://localhost:8000/api/redoc/

Exports are also available

• YAML: http://localhost:8000/api/docs/swagger.yaml
• JSON: http://localhost:8000/api/docs/swagger.yaml

Tests should be added in a tests package in each app, or in the root tests package for generic test cases.

Non-app-specific helper functions and classes should be added in the root libs package.

Kubernetes

For local development we recommend Minikube

1. Start the local Kubernetes Instance

    minikube start

2. Enable the registry: Documentation

    minikube addons enable registry

4. Forward the internal port of the registry to 127.0.0.1:5000 (Needs to stay open)

   kubectl port-forward -n kube-system replicationcontroller/registry 5000:5000

5. Enable the ingress

    minikube addons enable ingress

6. Optional: Patch hosts file to access via browser

    echo $(minikube ip) django-kompose.minikube.local django-kustomize.minikube.local django-helm.minikube.local | sudo tee -a /etc/hosts

Using Kompose

Conversion using environment defined in .env

    cd k8s/kompose
    docker-compose config | kompose convert --stdout -f - | kubectl apply -f -

Access: http://django-kompose.minikube.local

Using Kustomize

This repo contains a simple manifest for this application in k8s/kustomize/base There are two overlays available that you can have a look at

1. Minikube (Access: http://django-kustomize.minikube.local)

    cd k8s/kustomize
    kubectl apply -k overlays/minikube/

2. Production

    cd k8s/kustomize
    kubectl apply -k overlays/production/

Using helm (with helmfile)

This repo also contains a simple Helm chart for the application in k8s/helm/django To start it either use helm directly or run it using helmfile

    cd k8s/helmfile
    helmfile apply

Access: http://django-helm.minikube.local

Using Skaffold

On top of the three options there is also an option to use skaffold

    skaffold dev --port-forward