Terraform-EKS-Prometheous

About ℹ️

This repo contains the Terraform project to automate VPC + Elatic Kubernetes Cluster (EKS) deployment on AWS and setting up monitoring via Prometheus and Grafana stack using Helm Charts.

Key Componets 🧑‍💻

• vpc.tf : For creating Virtual Private Cloud in AWS
• eks.tf : For creating EKS Cluster
• terraform.tfvars : Terraform variables for VPC (use your own)
• k8s-config.yaml : Deployment for our Node App, Internal Service and ServiceMonitor that expose /metrics endpoint to Prometheus
• prometheus-node app : source code for our node application along with Dockerfile to build the image

Tech Used 💻

• Terraform
• Helm Charts
• Docker
• Kubernetes
• Prometheus
• Grafana

Part 1: Setting up VPC & EKS Custer in AWS ⚒️

Make sure you have Terraform installed on your machine and AWS CLI and Credentials configured. Navigate to the project root directory and follow the steps below"

Initialize Terraform

terraform init

Preview Terraform Actions

terraform plan

Apply configuration with variables

terraform apply -var-file terraform.tfvars

Destroy a Single Resource

terraform destroy -target aws_vpc.myapp-vpc

Destroy everything from tf files (clear up AWS resources)

terraform destroy

Show resources and components from current state

terraform state list

Show current state of a specific resource/data

terraform state show aws_vpc.myapp-vpc    

Set avail_zone as custom tf environment variable - before apply

export TF_VAR_avail_zone="ap-south-1"

For debugging:

export TF_LOG=TRACE    

---

Part 2: Setting up Promethous + Grafana Monitoring Stack

The folder prometheus-node contains a Node.js that utilizes package to expose metrics on /metrics endpoint and supports histogram, summaries, gauges and counters for the application stats.

Note: A public image for this node app is already created and available as 977977/node-prometheus. If you want to build and push it yourself, navigate to prometheus-node and run the following command:

docker build -t repo-name/image-name:image-tag .

docker push repo-name/image-name:image-tag

Deploy Prometheus Operator Stack

Make sure you have Helm installed on your machine.

Add Helm Chart Repo

helm repo add prometheus-community https://prometheus-community.github.io/helm-charts

helm repo update

Export EKS Kubeconfig to your environment

aws eks update-kubeconfig --name ava-eks-cluster --region ap-south-1

Create Namespace and install prometheus-grafana stack

kubectl create namespace monitoring-prom

helm install monitoring prometheus-community/kube-prometheus-stack -n monitoring-prom

helm ls -n monitoring-prom

Check Prometheus stack pods

kubectl get all -n monitoring-prom

Deploy Prometheous-Node Application and let ServiceMonnitor find it to scrap metrics

kubectl apply -f nginx-deployment-config.yaml

Access Prometheus UI

kubectl port-forward svc/monitoring-kube-prometheus-prometheus 9090:9090 -n monitoring-prom &

Access Grafana

kubectl port-forward svc/monitoring-grafana 8080:80 -n monitoring &
user: admin
pwd: prom-operator

Access the application

Go to EC2->LoadBalanancers->DNS. Click the endpoint to access Nginx server

or use:

kubectl port-forward svc/nodeappp 3000:3000

Contributions

Contributions are highly welcomed. Please send a Pull Request with suggested changes or open an Issue to get things started! Feel free to tag @sagar-uprety in the issues :)