Shorthub is a production-grade URL shortening platform demonstrating state-of-the-art DevOps practices. It features a fully automated CI/CD pipeline, canary deployments, and multi-environment orchestration on AWS EKS.
graph TD
subgraph "External"
UI_User[User Browser]
end
subgraph "AWS EKS Cluster"
LB[AWS ALB / Istio Gateway]
subgraph "prod Namespace"
pUI[UI Service - 5 Replicas]
pAPI[API Service - Canary Rollout]
pRedis[(Redis Cache)]
end
subgraph "staging Namespace"
sUI[UI Service]
sAPI[API Service]
end
end
subgraph "GitOps Pipeline"
GitHub[GitHub Repo]
GHA[GitHub Actions]
ArgoCD[ArgoCD]
ECR[(Amazon ECR)]
end
UI_User --> LB
LB --> pUI
pUI --> pAPI
pAPI --> pRedis
GitHub -- Push --> GHA
GHA -- Build/Push --> ECR
GHA -- Update Manifests --> GitHub
GitHub -- Sync --> ArgoCD
ArgoCD -- Deploy --> AWS_EKS
- Languages: Node.js 20 (Backend), React 18 (Frontend)
- Frameworks: Express, Vite, Tailwind CSS
- Infrastructure: Terraform, AWS EKS, VPC, IAM (IRSA)
- GitOps: ArgoCD, Argo Rollouts, Kustomize
- Observability: Prometheus, Grafana, Loki
- Service Mesh: Istio (Traffic Shifting)
- Zero-Touch Deployment: Push to
maintriggers image build, ECR push, and K8s manifest update. - Canary Strategy: 20% traffic shift to new API versions with automated health checks.
- Multi-Env Promotion: Environment-specific overlays (Dev/Staging/Prod) managed via ApplicationSets.
- Micro-Frontend Ready: React UI served via Nginx with reverse proxy to backend.
- High Availability: Multi-AZ node groups with Spot instance cost optimization.
- Clone the repo.
- Run
docker-compose up. - Open
http://localhost.
- Initialize Terraform:
terraform init && terraform apply. - Configure AWS Secrets in GitHub:
AWS_ACCESS_KEY_ID,AWS_SECRET_ACCESS_KEY. - Push to
mainbranch. - Access ArgoCD Dashboard:
kubectl get svc -n argocd.
- Architected a production-grade GitOps platform using EKS, VPC, and ArgoCD, reducing deployment manual intervention by 100%.
- Implemented advanced Canary deployment strategies with Argo Rollouts and Istio, enabling 20% traffic shifting and automated rollbacks based on 5xx error rate thresholds.
- Engineered a scalable multi-environment CI/CD pipeline with GitHub Actions and Kustomize, achieving zero-downtime deployments across Dev, Staging, and Prod.
- Automated Infrastructure as Code (IaC) with Terraform modules for modular VPC and EKS cluster management, utilizing IAM Roles for Service Accounts (IRSA) for least-privilege security.
- Enhanced system reliability by integrating Prometheus and Grafana for real-time application monitoring and automated health-based rollouts.