Production-grade multi-app EKS deployment

This project focuses on deploying a fully containerised 2048 game application on AWS EKS (Elastic Kubernetes Service). The infrastructure is provisioned using Terraform for consistency and scalability. CI/CD pipelines have been implemented using GitHub Actions to automate the deployment process, and monitoring tools like Prometheus and Grafana are configured to provide real-time insights into application performance and infrastructure health.

Tech Stack:

• Kubernetes: Orchestrates our containerised workloads for scalability and reliability.
• ArgoCD: Implementing GitOps for automated and declarative continuous delivery.
• Prometheus: Collects and queries app and infrastructure metrics for monitoring.
• Grafana: Visualises metrics and provides actionable insights through customised dashboards.
• Docker: Containerises the 2048 app to ensure a consistent runtime environment.

Architecture diagram:

Architecture

The deployment architecture is designed with scalability, reliability, and observability in mind. Here's a breakdown of the components:

1. Route 53 and ALB: DNS routing is handled via AWS Route 53, with HTTPS traffic secured using SSL certificates managed by cert-manager and terminated at an Application Load Balancer (ALB).
2. EKS Cluster: The application is hosted on an AWS EKS cluster, with worker nodes distributed across multiple availability zones for fault tolerance.
3. GitOps with ArgoCD: Continuous delivery is automated using ArgoCD, ensuring the Kubernetes cluster state reflects the desired state defined in Git repositories.
4. Terraform: Infrastructure provisioning and state management are fully automated using Terraform, with state locking enabled via DynamoDB and S3.
5. Monitoring: Prometheus collects detailed metrics, while Grafana visualises these metrics with customised dashboards for actionable insights.
6. CI/CD Pipeline: GitHub Actions automates application builds, containerisation (Docker), and deployments to EKS.

---

Features

1. Declarative Infrastructure:

   • EKS clusters, security groups, and VPCs are provisioned with Terraform, ensuring consistency across environments.
   • Utilises S3 for Terraform state storage and DynamoDB for state locking to avoid conflicts in collaborative workflows.

2. Automated Deployments:

   • GitHub Actions pipelines automate application builds and deployments.
   • Integration with ArgoCD ensures seamless GitOps workflows for continuous delivery.

3. Scalable and Secure Networking:

   • DNS routing via Route 53 and HTTPS termination at ALB ensures secure communication.
   • Kubernetes ingress handles internal routing to services.

4. Real-Time Monitoring:

   • Prometheus scrapes metrics from both the app and infrastructure components.
   • Grafana dashboards provide detailed visualisation for proactive monitoring.

---

Prometheus and Grafana Dashboard

Prometheus and Grafana are integrated into the deployment for real-time monitoring and observability of the Kubernetes cluster and applications.

• Prometheus: Used as the primary data collection and query backend. It scrapes metrics from Kubernetes components, application workloads, and infrastructure, providing detailed insights into resource usage and performance.
• Grafana: A visualisation tool layered on top of Prometheus, offering interactive dashboards and query options for monitoring cluster health, resource utilisation, and application performance.

Key Features:

• Cluster Monitoring: Dashboards include metrics like CPU utilisation, memory usage, and pod health across namespaces.
• Custom Dashboards: User-defined dashboards are set up for detailed application-specific monitoring.
• Alerting: Alerts can be configured in Prometheus and visualised in Grafana for proactive response to critical issues.

grafana dashboard

ArgoCD

ArgoCD is used to implement a GitOps workflow, ensuring that the Kubernetes cluster state is always in sync with the desired state defined in Git repositories.

Key Features:

• Automated Deployments: ArgoCD continuously monitors the Git repository for changes and applies them to the cluster.
• Declarative Configuration: All resources, including deployments, services, and ingress rules, are defined as code in Git.
• Application Health Monitoring: The ArgoCD dashboard provides real-time application health and sync status, ensuring transparency and traceability.