☁️ CloudSnap

A cloud-native image hosting platform, built to demonstrate an end-to-end DevOps pipeline — from containerization through orchestration, CI/CD, cloud deployment, and real-time monitoring.

Built as part of a research paper: "Design and Implementation of a Cloud-Native Application Deployment Framework Using Docker and Kubernetes"

---

🚀 What This Project Demonstrates

• Containerization — Flask app packaged with Docker
• Orchestration — 3 replicas running on Kubernetes (Minikube), behind a NodePort Service
• CI/CD — GitHub Actions automatically builds and pushes a new image to Docker Hub on every push
• Cloud Hosting — Live on AWS EC2 with a static Elastic IP
• Monitoring — Prometheus scrapes live metrics from every pod; Grafana visualizes request rate and pod uptime
• Resilience — Verified Kubernetes automatically recovers crashed pods without manual intervention

---

🏗️ Architecture

Developer → GitHub Repo → GitHub Actions (CI/CD) → Docker Hub
                                                        ↓
                                    ┌───────────────────┴───────────────────┐
                                    ↓                                       ↓
                    Kubernetes Cluster (Minikube)                  AWS EC2 Instance
                    3 Pod Replicas + NodePort Service              Docker Container + Elastic IP
                                    ↓
                    Prometheus (ServiceMonitor) → Grafana Dashboard

---

🧰 Tech Stack

Layer	Tool
Application	Flask (Python)
Containerization	Docker
Orchestration	Kubernetes (Minikube)
CI/CD	GitHub Actions
Cloud Hosting	AWS EC2 (Elastic IP)
Monitoring	Prometheus + Grafana
Metrics Export	prometheus_flask_exporter

---

📦 Getting Started on Your Own Machine

Prerequisites

Make sure you have these installed:

• Python 3.11+
• Docker Desktop
• Minikube
• kubectl
• Helm (for Prometheus/Grafana)
• Git

1. Clone the repository

git clone https://github.com/Midoriya-w/cloudSnap.git
cd cloudSnap

2. Run it locally (without Docker)

python -m venv .venv
# Windows
.venv\Scripts\Activate.ps1
# macOS/Linux
source .venv/bin/activate

pip install -r requirements.txt
python app.py

App will be running at http://localhost:5000 Metrics available at http://localhost:5000/metrics

3. Run it with Docker

docker build -t cloud-native-app .
docker run -d --name cloudsnap-app -p 5000:5000 cloud-native-app

Or pull the pre-built image directly:

docker pull midoriya543/cloud-native-app:latest
docker run -d --name cloudsnap-app -p 5000:5000 midoriya543/cloud-native-app:latest

4. Deploy to Kubernetes (Minikube)

minikube start
kubectl apply -f deployment.yaml
kubectl get pods       # confirm 3 pods are Running
kubectl get services    # confirm the NodePort service

minikube service cloud-native-app-service   # opens the app in your browser

5. Set up monitoring (Prometheus + Grafana)

helm repo add prometheus-community https://prometheus-community.github.io/helm-charts
helm repo update
helm install monitoring prometheus-community/kube-prometheus-stack

kubectl apply -f service-monitor.yaml

Get the Grafana admin password:

kubectl get secrets monitoring-grafana -o jsonpath="{.data.admin-password}" | base64 -d

Access Grafana:

kubectl port-forward svc/monitoring-grafana 3000:80

Then open http://localhost:3000 and log in with username admin and the password above.

---

🔁 CI/CD Pipeline

Every push to master triggers .github/workflows/ to:

1. Check out the repository
2. Build a new Docker image
3. Push it to Docker Hub as midoriya543/cloud-native-app:latest

No manual rebuild or redeploy needed — just push your code.

---

📊 Monitoring

• Prometheus scrapes /metrics from every pod every 15 seconds via a ServiceMonitor
• Grafana dashboard "CloudSnap Monitoring" tracks:
  • Request rate (rate(flask_http_request_total[5m]))
  • Pod uptime (up{job="cloud-native-app-service"})

---

📁 Project Structure

cloudSnap/
├── app.py                  # Flask application
├── Dockerfile               # Container build instructions
├── deployment.yaml          # Kubernetes Deployment + Service
├── service-monitor.yaml     # Prometheus ServiceMonitor
├── requirements.txt         # Python dependencies
├── frontend/                 # UI (HTML/CSS/JS)
├── .github/workflows/        # CI/CD pipeline
└── README.md

---

👥 Contributors

• Ch. Dinesh Sai Vardhan — Technical implementation
• Ch. S. M. Harsha — Documentation and literature review

Project guide: Dr. Sibendu Samanta, Assistant Professor, SRM University AP

---

📄 Research Paper

• Dockerize application
• Deploy on Kubernetes with 3 replicas
• Push image to Docker Hub
• GitHub Actions CI/CD pipeline
• Deploy on AWS EC2
• Prometheus + Grafana monitoring

---

🔮 Future Work

• NGINX Ingress for host/path-based routing and TLS termination
• Horizontal Pod Autoscaling driven by Prometheus metrics
• Automated testing in the CI/CD pipeline
• Grafana alerting for pod failures