☸️ Microservices Kubernetes Deployment -- Skill Test 2

Container Orchestration (MCAD)

---

Objective

Deploy a Node.js microservices application on Kubernetes, ensuring proper configuration, service discovery, and inter-service communication.

This deployment was performed on an on-prem Kubernetes cluster using Docker Desktop, not on EKS or any cloud provider.

---

Application Architecture

The application consists of four containerized Node.js microservices:

Service Name	Port
User Service	3000
Product Service	3001
Order Service	3002
Gateway Service	3003

---

Platform & Tools

• Kubernetes (Docker Desktop -- On-Prem)

• kubectl

• Docker

• NGINX Ingress Controller

---

Folder Structure

submission/
├── deployments/
│   ├── user-service.yaml
│   ├── product-service.yaml
│   ├── order-service.yaml
│   └── gateway-service.yaml
├── services/
│   ├── user-service.yaml
│   ├── product-service.yaml
│   ├── order-service.yaml
│   └── gateway-service.yaml
├── ingress/
│   └── ingress.yaml
├── screenshots/
│   ├── pods.png
│   ├── service-test.png
│   └── logs.png
├── user-service/
├── product-service/
├── order-service/
└── gateway-service/

---

Kubernetes Deployment Details

Deployments

Each Deployment includes:

• Correct container image reference

• Resource requests and limits

• Environment variables

• Liveness and readiness probes

• Proper labels and selectors

Services

• All services use ClusterIP

• Enables internal Kubernetes DNS-based service discovery

• Correct port mapping for each microservice

---

Deployment Steps

kubectl apply -f submission/deployments/ kubectl apply -f submission/services/ kubectl apply -f submission/ingress/ingress.yaml

---

Verification & Testing

1. Verify Pods, Services, and Ingress

kubectl get pods -o wide kubectl get svc kubectl get ingress

---

2. Inter-Service Communication Test (ClusterIP)

A temporary curl pod was used to validate Kubernetes DNS and service connectivity:

kubectl run curlpod --image=curlimages/curl -it --restart=Never -- sh

Inside the pod:

nslookup user-service curl -v http://user-service:3000

This confirms:

• Kubernetes DNS resolution

• ClusterIP service discovery

• TCP connectivity between services

Note: HTTP 404 Not Found responses are expected because the services do not expose a root (/) endpoint.
A 404 confirms that the service is reachable and the application is running.

---

3. Gateway Logs (Service Communication)

kubectl logs -l app=gateway-service --tail=100

Gateway logs confirm communication with downstream services.

---

Ingress Configuration (Bonus Task)

NGINX Ingress Controller was enabled and configured with the following routes:

Path	Backend Service
/api/users	User Service
/api/products	Product Service
/api/orders	Order Service
/	Gateway Service

Ingress Host:

micro.local

Ingress was tested using port-forwarding:

kubectl -n ingress-nginx port-forward svc/ingress-nginx-controller 8080:80

---

Screenshots

The following screenshots are included as evidence:

• pods.png -- Running pods (kubectl get pods -o wide)

• service-test.png -- DNS resolution and service connectivity test

• logs.png -- Gateway service logs showing communication

---

Troubleshooting

ImagePullBackOff

• Ensure Docker images are built locally when using Docker Desktop Kubernetes:

docker build -t user-service:latest ./submission/user-service

Curl Timeouts

• Some services do not expose / or /health endpoints.

• DNS resolution and TCP connectivity were used instead to validate service communication.

---

Conclusion

All microservices were successfully deployed on Kubernetes with:

• Proper service discovery

• Internal communication using ClusterIP

• Resource management and health probes

• Optional ingress routing

This fulfills all requirements of Skill Test 2 -- Container Orchestration.

📸 Screenshots (Attach in Submission)

---

🧑‍💻 Author

DEEPIKA NARENDRAN
DevOps Technical Lead | MCAD Program\

[DevOps B13] Skill Test 2 -- Cloud & Container Orchestration Microservices Kubernetes Deployment Assessment

📧 deepika2.ytb@gmail.com
💼 GitHub: JoinDeeHub
📍 Bengaluru, India