Introduction

Simple application demonstrate using Kubernetes in AWS with:

• EKS K8s Cluster
• Event Driven Architecture using NATs
• Monitoring and Login using Grafana, Loki, Prometheus through Helm charts.
• Gitops for K8s cluster using ArgoCD

Prerequisites:

• AWS CLI configured.
• Docker installed.
• Helm installed.
• ArgoCD installed on the cluster.
• kubectl installed and configured.
• eksctl installed.
• Minikube (locally + metric server + ingress addons are enabled)

Quick Start

1. EKS Cluster Setup:

• Create EKS cluster through AWS console.
• Create Node Group or Fargate Profile
• Create Role for cluster, Node Group.
• Create Access entry + policy for kubectl to access EKS cluster.

2. Deploy the ArgoCD Project:

   k kustomize argocd/overlays/{dev/prod} --enable-helm | kubectl apply -f -

2.1 Destroy the ArgoCD Project: bash k kustomize argocd/overlays/{dev/prod} --enable-helm | kubectl delete -f - # In case failed to delete application https://argo-cd.readthedocs.io/en/latest/user-guide/app_deletion/

3. Create local DNS (minikube):

   # File /etc/hosts
   127.0.0.1 simple-love-{dev/prod}.com
   127.0.0.1 simple-love-monitoring-{dev/prod}.com
   

   #Tunnel minikube cluster to localhost
   minikube tunnel

Access

1. ArgoCD UI

   minikube service argocd-server -n argocd
   # Get ArgoCD Password (User: admin)
   kubectl -n argocd get secret argocd-initial-admin-secret -o jsonpath="{.data.password}" | base64 --decode ; echo

2. Grafana UI

   kubectl get secret -n simple-love-monitoring-{dev/prod} monitoring-{dev/prod}-grafana -o jsonpath="{.data.admin-password}" | base64 --decode
   https://simple-love-monitoring-{dev/prod}.com

GitOps Workflow

Our GitOps workflow follows these principles:

1. Git as Single Source of Truth: All configuration is stored in this repository
2. Declarative Configuration: Desired state is declared in YAML files
3. Automated Synchronization: ArgoCD ensures the cluster state matches Git
4. Environment Promotion: Changes flow from dev → staging → production

Promotion Strategy

We use Kargo to implement a promotion strategy:

1. Development: Automatic deployment of new images
2. Staging: WIP
3. Production: Automatic deployment of new images

Usage Examples

View Deployment Status

kubectl get applications -n argocd

Setup Requirements

1. Add a Docker Hub token as a GitHub secret named DOCKER_HUB_TOKEN

2. Customize the workflow in .github/workflows/main.yml for additional microservices

Cleanup:

• Delete the Kubernetes EKS cluster and Helm related resources (ALB...)

Demo