wger-project-k8s

This is a working kubernetes realization of the wger-project/docker

Preparations

Nginx Docker Image

This project needs a Nginx image with a nginx.conf configured for the wger. So you can build it on your worker node.

cd ./wger-project-k8s/nginx/
docker build -t nginx-wger .

If you have multiple worker nodes it's better to push image to the Docker/private registry and than edit image's name in nginx-deployment.yaml to pull it.

Persistent Volumes

Inasmuch as this deploy was configured to use a NFS share you'll need to edit the persistentvolume.yaml file according to your NFS share's IP and folder structure or write own solution.

Deploy Wger to Kubernetes

kubectl apply -f ./wger-project-k8s/

It might take some time for the service to become available. Deploy will end when you'll see "Booting worker with pid: NNN" in the log of web pod. So don't panic when you will see "Done in NN s." but the error 502 will be still appearing.

Expose service on node port

To have access from the outside you can expose service on node's port

kubectl expose deployment nginx --type=NodePort --name=nginx-on-node-service

Default credentials

admin:adminadmin

Debug and diagnostic

Tools

• Kompose - Translate a Docker Compose File to Kubernetes Resources.

• LazyDocker - A simple terminal UI for both docker and docker-compose, written in Go with the gocui library.

• Dockly - Immersive terminal interface for managing docker containers and services written on NodeJS.

• K9S - Kubernetes CLI to manage your clusters in style.

Commands

# Attach terminal
kubectl exec --stdin --tty PODNAME -- /bin/bash

# Watch logs of pod
kubectl logs --follow PODNAME

# Get Services
kubectl get svc

# Check server's reply with redirections follow
curl -LI WorkerNodeIP:NodePort
curl -LI 10.96.191.134:8000

# Pod restart
kubectl delete pod PODNAME

Django Migrations

Django Migrations automatically starts, just wait for it. Fore some cases here's a hint how to do this manually in k8s. Place an actual web pod's name to the command and run it.

kubectl exec -it --namespace=default web-..... -- bash -c "python3 manage.py makemigrations && python3 manage.py migrate && python3 manage.py migrate --fake-initial && yarn install"

Delete and cleanup

# On master node
kubectl delete -f ./wger-project-k8s

# On worker node
docker system prune -a

# On NFS server
rm -rf /media/k8s/{data,media,postgres-data,static}/*