A web application for calculating Exponentially Weighted Moving Averages (EWMA) and simple averages. Built with FastAPI backend and React frontend. The application is containerized with Docker and deployed on a local K8s cluster using Kind.

├── backend/                 # FastAPI backend service
│   ├── app/
│   │   ├── main.py          # Application entry point
│   │   ├── routers/         # API route handlers
│   │   ├── schemas.py       # Pydantic models
│   │   └── services.py      # Business logic
│   ├── tests/               # Backend tests
│   ├── Dockerfile           # Backend container image
│   └── requirements.txt     # Python dependencies
│
├── frontend/                # Frontend application (React + Vite + TS)
│   ├── src/
│   │   ├── assets/          # Static assets (images, fonts, icons, etc.)
│   │   ├── components/      # Reusable UI components
│   │   ├── api/             # API calls 
│   │   ├── types/           # Shared types/interfaces
│   │   ├── App.tsx          # Root component
│   │   └── main.tsx         # Entry point
│   │
│   ├── public/              # Public static files
│   │   └── favicon.ico
│   │
│   ├── Dockerfile           # Frontend Docker image
│   ├── index.html           # Root HTML file
│   ├── package.json
│   ├── tsconfig.json
│   ├── tsconfig.app.json
│   ├── tsconfig.node.json
│   ├── vite.config.ts
│   ├── eslint.config.js
│   ├── .gitignore
│   └── README.md
│
├── k8s/                     # Kubernetes configurations
│   └── ewma-calculator/     # Helm chart
│       ├── templates/       # K8s resource templates
│       ├── values.yaml      # Default values
│       ├── values-staging.yaml
│       └── values-production.yaml
│
├── scripts/                 # Automation scripts
│   ├── build_images.sh      # Build Docker images
│   ├── push_images.sh       # Push images to Docker Hub
│   ├── test_backend.sh      # Run backend tests
│   ├── test_frontend.sh     # Test and build frontend
│   └── deploy.sh            # Deploy to Kubernetes (Kind in this case)
│
└── .github/workflows/       # CI/CD pipeline
  └── main.yml             # GitHub Actions workflow

• Docker
• kubectl
• Helm 3
• Kind (for local deployment)
• Python 3.12+ (for local development)
• Node.js 18+ (for local development)

cd backend
pip install -r requirements.txt
uvicorn app.main:app --reload

cd frontend
npm install
npm run dev

bash scripts/test_backend.sh

bash scripts/test_frontend.sh

bash scripts/build_images.sh

bash scripts/push_images.sh

In the root directory of this repo, there is a docker-compose.yaml file. Run the command below to locally run the application.

docker compose up

Frontend accessible at localhost:5173 and backend at localhost:8000

The GitHub Actions workflow automatically:

1. Runs backend tests
2. Runs frontend tests and builds
3. Builds Docker images with appropriate tags
4. Pushes images to Docker Hub

1. Push to main branch
2. Manual worklow dispatch

bash scripts/deploy.sh

This script will:

1. Install Kind and Helm (if not present)
2. Create a Kind cluster
3. Deploy staging environment
4. Deploy production environment
5. Set up port forwarding (for local testing)

1. Staging
   • Frontend: http://localhost:4000
   • Backend: http://localhost:9000
2. Production :
   • Frontend: http://localhost:3000
   • Backend: http://localhost:8000

kind create cluster --config k8s/ewma-calculator/kind-config.yaml --name ewma-cluster

cd k8s/ewma-calculator
helm upgrade --install ewma-staging . \
  --values values-staging.yaml \
  --wait

To see all running resources inside the staging namespace, run the command below

kubectl get all -n staging

helm upgrade --install ewma-production . \
  --values values-production.yaml \
  --wait

To see all running resources inside the production namespace, run the command below

kubectl get all -n production

Port forward to access locally

1. Frontend

kubectl port-forward -n staging svc/frontend-service 5173(any_port):80

2. Backend

kubectl port-forward -n staging svc/backend-service 8000(any_port):80

The same can be followed for production, just replace staging with production

curl http://localhost:8000/health

Should return: {"status":"healthy"}

Remove the Kind cluster

kind delete cluster --name ewma-cluster

🔎 Role of Metrics Server in HPA - Metrics Server provides live CPU and memory usage data from pods via the metrics.k8s.io API, which the Horizontal Pod Autoscaler (HPA) uses to scale workloads up or down.

⚠️ Important: Without Metrics Server, HPA cannot fetch CPU/memory usage.

Will use Helm to install metric server

1. Add Metrics Server repo

   helm repo add metrics-server https://kubernetes-sigs.github.io/metrics-server/

   This tells Helm where to find the metrics-server chart.
2. Update Helm repos

   helm repo update

   Ensures you have the latest charts from all configured repositories.
3. Skipping TLS verification

   helm up
   grade --install metrics-server metrics-server/metrics-server \ -n kube-system \ --set args={--kubelet-insecure-tls}

By default metric server uses TLS to connect to Kubelets. But since we are using a local environment Kind, kubelets don't have valid certificates, so we need the above command to bypass TLS verification