Microservices Application on AWS EKS

Here there is an example of a microservices-based application deployed on AWS EKS (Elastic Kubernetes Service).
It showcases containerization with Docker, orchestration with Kubernetes, and automatic scaling via Horizontal Pod Autoscaler (HPA).

---

Architecture Overview

The system is composed of three microservices:

Service	Description	Exposed Port
frontend	Simple web interface for user interaction	3000
gateway	Backend API gateway that routes requests to the NLP service	8080
nlp	Natural Language Processing microservice for text analysis	5000

Each service runs as a Docker container and is deployed on EKS using Kubernetes Deployments, Services, and HPAs.

---

Infrastructure Diagram

Client → [ LoadBalancer / Ingress ] → [ Frontend Service ]
↓
[ Gateway Service ]
↓
[ NLP Service ]

All components run inside the same Kubernetes cluster on EKS.
The Horizontal Pod Autoscaler (HPA) adjusts the number of pods per deployment based on CPU utilization.

---

Project Structure

.
├── k8s/
│   ├── deployments/
│   │   ├── frontend-deploy.yaml
│   │   ├── gateway-deploy.yaml
│   │   └── nlp-deploy.yaml
│   ├── services/
│   │   ├── frontend-svc.yaml
│   │   ├── gateway-svc.yaml
│   │   └── nlp-svc.yaml
│   ├── autoscaling/
│   │   └── hpa.yaml
│   └── ingress/
│       └── ingress.yaml
├── docker/
│   ├── frontend/Dockerfile
│   ├── gateway/Dockerfile
│   └── nlp/Dockerfile
├── src/
│   ├── frontend/
│   ├── gateway/
│   └── nlp/
└── README.md

---

Deployment Instructions

1. Build Docker Images

Each microservice includes its own Dockerfile.
Build and push them to your Amazon ECR repository:

aws ecr create-repository --repository-name frontend
aws ecr create-repository --repository-name gateway
aws ecr create-repository --repository-name nlp

# Example build and push
docker build -t <aws_account_id>.dkr.ecr.<region>.amazonaws.com/frontend:latest ./docker/frontend
docker push <aws_account_id>.dkr.ecr.<region>.amazonaws.com/frontend:latest

2. Apply Kubernetes Manifests

Once images are available in ECR, deploy everything to your EKS cluster:

kubectl apply -f k8s/deployments/
kubectl apply -f k8s/services/
kubectl apply -f k8s/autoscaling/hpa.yaml
kubectl apply -f k8s/ingress/

3. Verify Deployment

Check that pods, services, and HPAs are running:

kubectl get pods
kubectl get svc
kubectl get hpa

---

Autoscaling (HPA)

The file k8s/autoscaling/hpa.yaml defines autoscaling policies for all microservices:

Service	Min Pods	Max Pods	Target CPU Utilization
frontend	2	6	60%
gateway	2	10	60%
nlp	2	8	70%

Ensure the Metrics Server is installed on your cluster:

kubectl apply -f https://github.com/kubernetes-sigs/metrics-server/releases/latest/download/components.yaml

---

Technologies Used

• AWS EKS – Managed Kubernetes cluster
• Kubernetes – Deployments, Services, Ingress, Autoscaling
• Docker – Containerization of each microservice
• Horizontal Pod Autoscaler (HPA) – Dynamic scaling based on CPU usage
• Amazon ECR – Private container registry
• (Optional) Ingress Controller (NGINX / ALB) – Routing to frontend

---

---

Environment Variables & Secrets

Before deploying the application or using CI/CD pipelines, make sure to configure the following secrets and environment variables in your GitHub repository (or local .env files).

AWS & EKS Configuration

Variable	Description	Example
AWS_ACCESS_KEY_ID	AWS access key for programmatic access	AKIAIOSFODNN7EXAMPLE
AWS_SECRET_ACCESS_KEY	AWS secret access key	wJalrXUtnFEMI/K7MDENG/bPxRfiCYEXAMPLEKEY
AWS_REGION	AWS region where your EKS cluster and ECR repositories are deployed	eu-central-1
EKS_CLUSTER_NAME	Name of the target EKS cluster	eks-ml-api
ECR_REGISTRY	ECR registry URL used for Docker image pushes	<aws_account_id>.dkr.ecr.<region>.amazonaws.com

---

GitHub Actions (CI/CD)

If you are using GitHub Actions for automated deployment or image build & push, define these in your repository under
Settings → Secrets and variables → Actions → New repository secret:

Secret	Description
AWS_ACCESS_KEY_ID	AWS programmatic access key
AWS_SECRET_ACCESS_KEY	AWS secret access key
EKS_CLUSTER_NAME	Target cluster name used in deployment jobs
AWS_REGION	AWS region
ECR_REGISTRY	Your ECR registry endpoint

---

💡 Local Development (Optional)

If you run the services locally using docker-compose or minikube, create a .env file at the project root:

AWS_REGION=eu-central-1
EKS_CLUSTER_NAME=eks-ml-api
ECR_REGISTRY=123456789012.dkr.ecr.eu-central-1.amazonaws.com
FRONTEND_IMAGE=${ECR_REGISTRY}/frontend:latest
GATEWAY_IMAGE=${ECR_REGISTRY}/gateway:latest
NLP_IMAGE=${ECR_REGISTRY}/nlp:latest

---

Future Improvements

• Add Prometheus + Grafana for monitoring
• Implement autoscaling based on memory or custom metrics
• Add CI/CD pipeline with GitHub Actions
• Add persistent storage for the NLP service

---

Author

Tommaso Fiorillo Cloud & DevOps Enthusiast ☁️ GitHub • LinkedIn