Final Project – Deploying a Multi-Service App

Bring together everything you've learned to deploy a real, multi-service application on your K3s cluster with persistence, ingress, and CI/CD integration.

🎉 You’ve made it to the final project!

Now it’s time to apply everything you’ve learned and build a real Kubernetes deployment on your Raspberry Pi cluster using K3s.

🎯 Objective

Deploy a complete multi-service web application with:

• A frontend (e.g., Nginx or Flask)
• A backend API (e.g., FastAPI or Node.js)
• A Redis or PostgreSQL database
• Persistent storage
• Ingress routing with TLS
• Optional: CI/CD integration via GitHub Actions or ArgoCD

📁 Project Structure Example

Your repo should include:

k8s/
├── frontend-deployment.yaml
├── backend-deployment.yaml
├── database-deployment.yaml
├── services.yaml
├── ingress.yaml
├── pvc.yaml
└── secrets.yaml

✅ Requirements

• All services deployed with Kubernetes Deployments
• Internal Services defined with ClusterIP
• One or more PersistentVolumeClaims for data
• Public access via Ingress
• Optionally: Helm chart or GitOps integration

🛠 Steps to Complete

1. Choose Your Stack

You can build your own or fork an existing sample, such as:

• Flask + Redis
• React + FastAPI + PostgreSQL
• WordPress + MariaDB (via Helm)

2. Write Deployment and Service YAMLs

Example:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: api
spec:
  replicas: 1
  template:
    spec:
      containers:
        - name: api
          image: kevdev/my-api:latest
          ports:
            - containerPort: 8000
---
apiVersion: v1
kind: Service
metadata:
  name: api
spec:
  ports:
    - port: 80
      targetPort: 8000
  selector:
    app: api

3. Set Up Storage and Secrets

• Create a PVC for databases
• Store credentials using kubectl create secret generic ...

4. Configure Ingress and TLS

Use a subdomain like app.local with an Ingress rule and test via /etc/hosts.

5. Automate Deployment

• Use kubectl apply -f k8s/ via GitHub Actions
• Or sync a Git repo with ArgoCD or Flux

🧪 Bonus Challenges

• Enable horizontal pod autoscaling
• Add Prometheus + Grafana monitoring
• Use a custom Helm chart to package your app
• Deploy to a second K3s cluster (multi-cluster practice)

📸 Share Your Work

• Take a photo of your Raspberry Pi cluster
• Share your architecture diagram or GitHub repo
• Post your build on GitHub, Mastodon, or Reddit with the tag #k3sOnPi

🏁 Summary

This final project ties together:

• Node setup
• K3s cluster installation
• Deployments, services, storage, ingress, and GitOps

You’ve built a full-featured Kubernetes platform on affordable hardware — congrats! 🎉

Next up: Course Summary

< Previous Next >

You can use the arrows  ← → on your keyboard to navigate between lessons.