After 7 months of DevOps training, students should build something real, production-style, enterprise-level — not just deploy a container.

🎯 PROJECT GOAL

Students must build and deploy a 6-microservice cloud-native application using full CI/CD + GitOps + Monitoring + Infrastructure as Code.

They must:

• Provision infrastructure with Terraform
• Build Docker images
• Push to ECR
• Create CI pipeline with Jenkins
• Deploy to EKS using Helm
• Use Argo CD for GitOps
• Store secrets in AWS Secrets Manager
• Monitor with Prometheus + Grafana
• Deploy one microservice to ECS (to show orchestration comparison)

🏗️ ARCHITECTURE OVERVIEW

Developer → GitHub
        ↓
      Jenkins (CI)
        ↓
      Docker build
        ↓
      Push to ECR
        ↓
     Argo CD watches GitOps repo
        ↓
      Helm deploys to EKS
        ↓
   Prometheus scrapes metrics
        ↓
      Grafana dashboards

Separate:

• 1 microservice deployed to ECS (Fargate)
• 5 microservices deployed to EKS

📦 APPLICATION STRUCTURE

Students must create:

6 Microservices

1. user-service
2. order-service
3. payment-service
4. product-service
5. notification-service
6. gateway-service (API gateway / ingress)

Language: Node.js / Python / Java (your choice)

Each must:

• Have its own Dockerfile
• Expose REST API
• Have health endpoint /health
• Expose /metrics endpoint (Prometheus format)

📁 GITHUB REPOSITORY STRUCTURE

They must create 3 repositories:

1️⃣ app-repo (Source Code)

microservices-platform/
 ├── user-service/
 ├── order-service/
 ├── payment-service/
 ├── product-service/
 ├── notification-service/
 └── gateway-service/

2️⃣ helm-charts-repo

helm-charts/
 ├── user-service/
 ├── order-service/
 ├── payment-service/
 ├── product-service/
 ├── notification-service/
 └── gateway-service/

Each chart must include:

• Deployment.yaml
• Service.yaml
• values.yaml
• HPA.yaml
• Ingress.yaml (for gateway)

3️⃣ gitops-repo

gitops/
 ├── dev/
 │    ├── user.yaml
 │    ├── order.yaml
 │    └── ...
 └── prod/

Argo CD watches this repo.

🧱 INFRASTRUCTURE TASK (Terraform)

create Terraform project:

terraform/
 ├── vpc.tf
 ├── eks.tf
 ├── ecs.tf
 ├── ecr.tf
 ├── iam.tf
 ├── secrets.tf
 ├── monitoring.tf
 └── outputs.tf

Terraform must provision:

• VPC (public + private subnets)
• EKS cluster
• ECS cluster (Fargate)
• 6 ECR repositories
• IAM roles for EKS & ECS
• AWS Secrets Manager secret
• Security groups
• ALB
• Route53 record (optional bonus)

🔐 SECRETS MANAGEMENT

• Store DB password in AWS Secrets Manager
• Retrieve it in:

• EKS using External Secrets Operator
• ECS using task definition secret reference

They must NOT hardcode passwords.

🚀 CI TASK (JENKINS)

Multi-branch pipeline

Pipeline stages:

1. Checkout
2. Unit Test
3. Build Docker image
4. Tag with Git SHA
5. Push to ECR
6. Update GitOps repo image tag
7. Commit & push to GitOps repo

Bonus:

• Add SonarQube scan
• Add Trivy security scan

🔁 CD TASK (ARGO CD)

• Install Argo CD in EKS
• Connect to GitOps repo
• Create Application CRDs
• Enable auto-sync
• Enable self-heal

must demonstrate:

• Changing image tag in GitOps repo
• Argo automatically deploys new version

📊 MONITORING TASK

Install via Helm:

• Prometheus
• Grafana
• kube-state-metrics
• node-exporter

must:

• Expose metrics endpoint in microservices
• Configure ServiceMonitor

• Create Grafana dashboard:

  • Pod CPU
  • Memory
  • Request rate
  • Error rate

🐳 ECS TASK (Comparison)

Deploy payment-service to:

• ECS Fargate
• Behind ALB

explain difference between:

EKS vs ECS:

• Control plane
• Scaling
• Cost
• Flexibility

📈 SCALING TASK

• Create HPA for at least 2 services
• Demonstrate load test
• Show pods scale up

Bonus:

• Create cluster autoscaler

🔎 TROUBLESHOOTING SCENARIOS

random failure scenarios:

1. ImagePullBackOff
2. CrashLoopBackOff
3. Secret not injected
4. Prometheus not scraping
5. ALB health check failing
6. Terraform state lock

You have to able to debug live.

📋 FINAL PRESENTATION REQUIREMENT

Each student must explain:

• Architecture diagram
• CI flow
• CD flow
• GitOps model
• Secrets handling
• Monitoring setup
• Scaling behavior
• How rollback works

🎓 EVALUATION CRITERIA

🧠 WHAT THIS PROJECT PROVES

• Mid-level DevOps
• Platform Engineer
• Cloud DevOps
• Kubernetes Engineer

This is real enterprise level.