Kubernetes Multi-Cloud Deployment with Environments

This repository provides a Kubernetes multi-cloud deployment setup with three environments: dev, staging, and prod. It uses ArgoCD for GitOps, Prometheus for monitoring, and Grafana for visualization. The architecture supports deployment on a single cloud provider or a redundant setup across multiple clouds (e.g., AWS, GCP, Azure).

Features

• Three Environments: Separate namespaces for dev, staging, and prod with tailored resource configurations.
• GitOps with ArgoCD: Automates deployment and synchronization of Kubernetes manifests.
• Monitoring with Prometheus: Collects metrics from each environment.
• Visualization with Grafana: Provides dashboards for monitoring data, accessible per environment.
• Multi-Cloud Support: Deploy on one or multiple Kubernetes clusters across different cloud providers.
• Extensible: Add your application manifests in manifests/environments/<env>/.

Prerequisites

• Kubernetes clusters (e.g., AWS EKS, GCP GKE, Azure AKS) with kubectl configured.
• kubectl and kubeconfig set up with contexts for each cluster (e.g., dev-cluster, staging-cluster, prod-cluster).
• Helm (optional, for manual chart installations).
• GitHub repository access to store this project.

Setup Instructions

1. Clone the Repository

git clone https://github.com/alex-dembele/k8s-multi-cloud-deployment.git

cd k8s-multi-cloud-deployment

2. Configure Kubernetes Clusters

Update scripts/setup-clusters.sh with your cluster details (e.g., AWS EKS, GCP GKE, Azure AKS). Example for AWS EKS:

chmod +x scripts/setup-clusters.sh ./scripts/setup-clusters.sh

Verify contexts:

kubectl config get-contexts

3. Update Repository URL

Replace in manifests/applications/*.yaml with your GitHub username.

4. Deploy to a Cluster

Run the deployment script for a specific cluster and environment:

chmod +x scripts/apply-configs.sh ./scripts/apply-configs.sh

Example: ./scripts/apply-configs.sh dev-cluster dev

Repeat for other environments:

./scripts/apply-configs.sh staging-cluster staging ./scripts/apply-configs.sh prod-cluster prod

Multi-Cloud Deployment

This repository includes helper scripts to deploy the same environment across multiple cluster contexts (cloud providers).

• Make the scripts executable:

chmod +x scripts/setup-clusters.sh scripts/apply-configs.sh scripts/deploy-multi-cloud.sh

• Configure your clusters (examples provided for AWS/GCP/Azure):

./scripts/setup-clusters.sh aws gcp azure
# or run without args to execute the AWS example
./scripts/setup-clusters.sh

• Deploy one environment to multiple clusters in one command:

./scripts/deploy-multi-cloud.sh prod aws-prod-cluster gcp-prod-cluster azure-prod-cluster

If you want ArgoCD to point to a different Git repository for manifests, apply-configs.sh accepts an optional third parameter repo-url that will patch the ArgoCD Application manifests on-the-fly.

For a full "bootstrap" walkthrough including cluster provisioning examples and a sample application, see docs/bootstrap.md.

5. Access ArgoCD

Get the ArgoCD server URL for the cluster:

kubectl -n argocd get svc argocd-server

Access ArgoCD UI via the LoadBalancer IP. Retrieve the admin password:

kubectl -n argocd get secret argocd-initial-admin-secret -o jsonpath="{.data.password}" | base64 -d

6. Access Grafana

Get the Grafana service URL for the environment:

kubectl -n get svc grafana

Example: kubectl -n dev get svc grafana Access Grafana via the LoadBalancer IP. Default credentials: admin/admin.

7. Multi-Cloud Redundancy

To deploy across multiple clouds for an environment: Configure additional clusters in scripts/setup-clusters.sh (e.g., GCP or Azure for prod). Run apply-configs.sh for each cluster context and environment:

./scripts/apply-configs.sh gcp-prod-cluster prod ./scripts/apply-configs.sh azure-prod-cluster prod

ArgoCD will sync the same manifests across all clusters, ensuring consistency.

Environment Configurations Dev: Lower resource limits (e.g., 512Mi memory for Prometheus) for development. Staging: Moderate resources to mimic production but with less scale. Prod: Higher resources (e.g., 2Gi memory for Prometheus) for production reliability.

Extending the Project

Add application manifests in manifests/environments//.

Create new ArgoCD Application resources in manifests/applications/ to manage them.

Customize Grafana dashboards or Prometheus ServiceMonitors in manifests/environments//.

Troubleshooting

ArgoCD Sync Issues: Check the ArgoCD UI for sync status and logs.

Prometheus/Grafana Issues: Verify ServiceMonitors and Pod status in the environment namespace:

kubectl -n get pods

Cluster Access: Ensure kubectl contexts are correctly configured.

Contributing

Feel free to open issues or submit pull requests to improve this project.

.gitignore

Kubernetes *.kubeconfig Local env files .env *.local