terraform-aws-eks

Quickstart

Deploy your first EKS cluster in minutes with a working Terraform configuration.

Compute Resources

Choose between EKS Auto Mode, managed node groups, self-managed nodes, and Fargate.

Configuration

Configure networking, access control, add-ons, IRSA, and encryption.

Reference

Complete reference for all module inputs and outputs.

What is terraform-aws-eks?

terraform-aws-eks is a community-maintained Terraform module that creates production-ready Amazon EKS clusters. It wraps the AWS EKS service with opinionated defaults and exposes a rich set of configuration options — so you get a secure, well-configured cluster without writing hundreds of lines of Terraform from scratch. The module handles the full lifecycle of an EKS cluster:

Control plane — EKS cluster, IAM roles, KMS encryption, CloudWatch logging, security groups
Data plane — EKS managed node groups, self-managed node groups, Fargate profiles, EKS Auto Mode
Access — Cluster access entries, IRSA (IAM Roles for Service Accounts), OIDC provider
Add-ons — CoreDNS, kube-proxy, VPC CNI, EKS Pod Identity Agent, and any community add-on

terraform-aws-eks usage

module "eks" {
  source  = "terraform-aws-modules/eks/aws"
  version = "~> 21.0"

  name               = "my-cluster"
  kubernetes_version = "1.33"

  vpc_id     = "vpc-1234556abcdef"
  subnet_ids = ["subnet-abcde012", "subnet-bcde012a", "subnet-fghi345a"]

  # EKS Auto Mode — fully managed data plane
  compute_config = {
    enabled    = true
    node_pools = ["general-purpose"]
  }

  enable_cluster_creator_admin_permissions = true

  tags = {
    Environment = "production"
    Terraform   = "true"
  }
}

Key features

EKS Auto Mode

Fully managed data plane with AWS-controlled node pools. No node group management required.

Managed Node Groups

AWS-managed EC2 node groups with custom launch templates, AMI types, and auto-repair.

Self-Managed Nodes

Full control over EC2 Auto Scaling groups, mixed instance policies, and lifecycle hooks.

Fargate Profiles

Serverless compute — run pods without managing EC2 instances.

Karpenter

Intelligent, just-in-time node provisioning via the included Karpenter sub-module.

EKS Hybrid Nodes

Connect on-premises and edge nodes to your EKS cluster over SSM or IAM Roles Anywhere.

IRSA

OIDC-based IAM Roles for Service Accounts — fine-grained pod-level AWS permissions.

Cluster Access Entries

API-based IAM authentication. Supports STANDARD, EC2_LINUX, FARGATE_LINUX, and HYBRID_LINUX types.

KMS Encryption

Automatic KMS key creation and secret encryption with configurable key rotation.

Compute options at a glance

EKS Auto Mode
Managed Node Groups
Self-Managed Node Groups
Fargate

AWS manages the entire data plane. You define node pool types; AWS handles provisioning, scaling, patching, and termination.

compute_config = {
  enabled    = true
  node_pools = ["general-purpose", "system"]
}

Best for: Teams that want minimal operational overhead and are comfortable with AWS-managed nodes.

AWS manages the EC2 instances within the node group, but you control instance types, AMIs, and scaling parameters.

eks_managed_node_groups = {
  main = {
    ami_type       = "AL2023_x86_64_STANDARD"
    instance_types = ["m5.xlarge"]
    min_size       = 2
    max_size       = 10
    desired_size   = 2
  }
}

Best for: Most production workloads needing control over instance types and AMIs.

You own the EC2 Auto Scaling group. Full control over every launch template parameter.

self_managed_node_groups = {
  main = {
    instance_type = "m5.xlarge"
    min_size      = 2
    max_size      = 10
    desired_size  = 2
  }
}

Best for: Advanced use cases requiring mixed instance policies, custom lifecycle hooks, or placement groups.

Serverless — each pod runs in its own isolated compute environment. No nodes to manage.

fargate_profiles = {
  default = {
    selectors = [
      { namespace = "kube-system" },
      { namespace = "default" }
    ]
  }
}

Best for: Batch workloads, dev/test environments, or workloads that don’t need persistent local storage.

Requirements

Requirement	Version
Terraform	`>= 1.5.7`
AWS Provider	`>= 6.28`
time Provider	`>= 0.9`
tls Provider	`>= 4.0`

Getting started

Add the module to your Terraform configuration

Reference the module from the Terraform Registry and pin to a version:

module "eks" {
  source  = "terraform-aws-modules/eks/aws"
  version = "~> 21.0"
}

Configure your cluster

Set the required variables — cluster name, Kubernetes version, VPC, and subnets. Choose your compute strategy.

Initialize and apply

terraform init
terraform plan
terraform apply

Access your cluster

aws eks update-kubeconfig --name my-cluster --region us-east-1
kubectl get nodes

Follow the full quickstart guide

Walk through a complete working example from zero to a running EKS cluster.

Get Started

Compute Resources

Configuration

Sub-Modules

Examples

Operations

Quickstart

Compute Resources

Configuration

Reference

What is terraform-aws-eks?

Key features

EKS Auto Mode

Managed Node Groups

Self-Managed Nodes

Fargate Profiles

Karpenter

EKS Hybrid Nodes

IRSA

Cluster Access Entries

KMS Encryption

Compute options at a glance

Requirements

Getting started

Follow the full quickstart guide

Build docs developers (and LLMs) love

Get Started

Compute Resources

Configuration

Sub-Modules

Examples

Operations

Quickstart

Compute Resources

Configuration

Reference

​What is terraform-aws-eks?

​Key features

EKS Auto Mode

Managed Node Groups

Self-Managed Nodes

Fargate Profiles

Karpenter

EKS Hybrid Nodes

IRSA

Cluster Access Entries

KMS Encryption

​Compute options at a glance

​Requirements

​Getting started

Follow the full quickstart guide

Build docs developers (and LLMs) love

What is terraform-aws-eks?

Key features

Compute options at a glance

Requirements

Getting started