gym-tracker/kubernetes

Kubernetes Deployment

This directory contains Kubernetes manifests for deploying the Gym Tracker application.

Prerequisites

• Kubernetes cluster (1.19+)
• kubectl configured to access your cluster
• Docker registry (optional, for remote deployments)

Files

• persistentvolumeclaim.yaml - PVC for SQLite database storage
• deployment.yaml - Application deployment
• service.yaml - ClusterIP service for internal access
• ingress.yaml - Optional ingress for external access

Deployment Steps

1. Build and Push Docker Image

Build the Docker image:

docker build -t gym-tracker:latest .

If deploying to a remote cluster, tag and push to your registry:

docker tag gym-tracker:latest your-registry/gym-tracker:latest
docker push your-registry/gym-tracker:latest

Update the image name in deployment.yaml accordingly.

2. Deploy to Kubernetes

Apply all manifests:

kubectl apply -f kubernetes/

Or apply individually in order:

kubectl apply -f kubernetes/persistentvolumeclaim.yaml
kubectl apply -f kubernetes/deployment.yaml
kubectl apply -f kubernetes/service.yaml
# Optional: kubectl apply -f kubernetes/ingress.yaml

3. Verify Deployment

Check the deployment status:

kubectl get pods -l app=gym-tracker
kubectl get svc gym-tracker
kubectl get pvc gym-tracker-data

View logs:

kubectl logs -l app=gym-tracker -f

4. Access the Application

Port Forward (for testing)

kubectl port-forward svc/gym-tracker 8080:80

Then access at http://localhost:8080

Using Ingress (for production)

1. Ensure an Ingress controller is installed in your cluster
2. Update the host in ingress.yaml with your domain
3. Apply the ingress: kubectl apply -f kubernetes/ingress.yaml
4. Access at http://your-domain.com

Using LoadBalancer

Change the service type in service.yaml from ClusterIP to LoadBalancer:

spec:
  type: LoadBalancer

Then get the external IP:

kubectl get svc gym-tracker

Configuration

Storage

The default PVC requests 1Gi of storage. Adjust in persistentvolumeclaim.yaml:

resources:
  requests:
    storage: 5Gi  # Increase as needed

Replicas

The application uses SQLite, which is file-based. Keep replicas at 1 to avoid database locking issues:

spec:
  replicas: 1

For high availability, consider migrating to PostgreSQL or MySQL.

Resources

Adjust resource limits in deployment.yaml based on your needs:

resources:
  requests:
    memory: "256Mi"
    cpu: "200m"
  limits:
    memory: "1Gi"
    cpu: "1000m"

Backup

To backup the SQLite database:

kubectl exec -it <pod-name> -- sqlite3 /app/data/gym-tracker.db ".backup /app/data/backup.db"
kubectl cp <pod-name>:/app/data/backup.db ./backup.db

Troubleshooting

Pod not starting

kubectl describe pod -l app=gym-tracker
kubectl logs -l app=gym-tracker

Database issues

Check volume mount:

kubectl exec -it <pod-name> -- ls -la /app/data

Health check failures

Test health endpoint:

kubectl exec -it <pod-name> -- wget -O- http://localhost/health

Cleanup

Remove all resources:

kubectl delete -f kubernetes/

Warning: This will delete the PVC and all data. Backup first if needed.