Skip to main content
Sponsor @cloudposse

Module: eks-node-group

Terraform module to provision an EKS Managed Node Group for Elastic Kubernetes Service.

Instantiate it multiple times to create EKS Managed Node Groups with specific settings such as GPUs, EC2 instance types, or autoscale parameters.

IMPORTANT: When SSH access is enabled without specifying a source security group, this module provisions EKS Node Group nodes that are globally accessible by SSH (22) port. Normally, AWS recommends that no security group allows unrestricted ingress access to port 22 .

Introduction​

This module creates an EKS Managed Node Group for an EKS cluster. It assumes you have already created an EKS cluster, but you can create the cluster and the node group in the same Terraform configuration. See our full-featured root module (a.k.a. component) eks/cluster for an example of how to do that.

Launch Templates​

This module always uses a launch template to create the node group. You can create your own launch template and pass in its ID, or else this module will create one for you.

The AWS default for EKS is that if the launch template is updated, the existing nodes will not be affected. Only new instances added to the node group would get the changes specified in the new launch template. In contrast, when the launch template changes, this module can immediately create a new node group from the new launch template to replace the old one.

See the inputs create_before_destroy and immediately_apply_lt_changes for details about how to control this behavior.

Operating system differences​

Currently, EKS supports 4 Operating Systems: Amazon Linux 2, Amazon Linux 2023, Bottlerocket, and Windows Server. This module supports all 4 OSes, but support for detailed configuration of the nodes varies by OS. The 4 inputs:

  1. before_cluster_joining_userdata
  2. kubelet_additional_options
  3. bootstrap_additional_options
  4. after_cluster_joining_userdata

are fully supported for Amazon Linux 2 and Windows, and take advantage of the bootstrap.sh supplied on those AMIs. NONE of these inputs are supported on Bottlerocket. On AL2023, only the first 2 are supported.

Note that for all OSes, you can supply the complete userdata contents, which will be untouched by this module, via userdata_override_base64.

Usage​

Major Changes (breaking and otherwise)​

With the v3.0.0 release of this module, support for Amazon Linux 2023 (AL2023) has been added, and some breaking changes have been made. Please see the release notes for details.

With the v2.0.0 (a.k.a. v0.25.0) release of this module, it has undergone major breaking changes and added new features. Please see the migration document for details.

For a complete example, see examples/complete.

For automated tests of the complete example using bats and Terratest (which tests and deploys the example on AWS), see test.

Sources of Information​

  • The code examples below are manually updated and have a tendency to fall out of sync with actual code, particularly with respect to usage of other modules. Do not rely on them.
  • The documentation on this page about this module's inputs, outputs, and compliance is all automatically generated and is up-to-date as of the release date. After the code itself, this is your best source of information.
  • The code in examples/complete is automatically tested before every release, so that is a good place to look for verified example code. Keep in mind, however, it is code for testing, so it may not represent average use cases or best practices.
  • Of course, the READMEs and examples/complete directories in the other modules' GitHub repos are more authoritative with respect to how to use those modules than this README is.

Example Code​

provider "aws" {
  region = var.region
}

module "label" {
  source = "cloudposse/label/null"
  # Cloud Posse recommends pinning every module to a specific version
  # version  = "x.x.x"

  namespace  = var.namespace
  name       = var.name
  stage      = var.stage
  delimiter  = var.delimiter
  attributes = ["cluster"]
  tags       = var.tags
}

locals {
  # Prior to Kubernetes 1.19, the usage of the specific kubernetes.io/cluster/* resource tags below are required
  # for EKS and Kubernetes to discover and manage networking resources
  # https://www.terraform.io/docs/providers/aws/guides/eks-getting-started.html#base-vpc-networking
  tags = { "kubernetes.io/cluster/${module.label.id}" = "shared" }
}

module "vpc" {
  source = "cloudposse/vpc/aws"
  # Cloud Posse recommends pinning every module to a specific version
  # version = "1.x.x"

  cidr_block = "172.16.0.0/16"

  tags    = local.tags
  context = module.label.context
}

module "subnets" {
  source = "cloudposse/dynamic-subnets/aws"
  # Cloud Posse recommends pinning every module to a specific version
  # version = "2.x.x"

  availability_zones   = var.availability_zones
  vpc_id               = module.vpc.vpc_id
  igw_id               = [module.vpc.igw_id]
  ipv4_cidr_block      = [module.vpc.vpc_cidr_block]
  nat_gateway_enabled  = true
  nat_instance_enabled = false

  tags    = local.tags
  context = module.label.context
}

module "eks_cluster" {
  source = "cloudposse/eks-cluster/aws"
  # Cloud Posse recommends pinning every module to a specific version
  # version = "4.x.x"

  vpc_id     = module.vpc.vpc_id
  subnet_ids = module.subnets.public_subnet_ids

  kubernetes_version    = var.kubernetes_version
  oidc_provider_enabled = true

  context = module.label.context
}

module "eks_node_group" {
  source = "cloudposse/eks-node-group/aws"
  # Cloud Posse recommends pinning every module to a specific version
  # version     = "3.x.x"

  instance_types        = [var.instance_type]
  subnet_ids            = module.subnets.public_subnet_ids
  min_size              = var.min_size
  max_size              = var.max_size
  cluster_name          = module.eks_cluster.eks_cluster_id
  create_before_destroy = true
  kubernetes_version    = var.kubernetes_version == null || var.kubernetes_version == "" ? [] : [var.kubernetes_version]

  # Enable the Kubernetes cluster auto-scaler to find the auto-scaling group
  cluster_autoscaler_enabled = var.autoscaling_policies_enabled

  context = module.label.context

  # Ensure the cluster is fully created before trying to add the node group
  module_depends_on = [module.eks_cluster.kubernetes_config_map_id]
}

Windows Managed Node groups​

Windows managed node-groups have a few pre-requisites.

  • Your cluster must contain at least one linux based worker node
  • Your EKS Cluster must have the AmazonEKSVPCResourceController and AmazonEKSClusterPolicy policies attached
  • Your cluster must have a config-map called amazon-vpc-cni with the following content
apiVersion: v1
kind: ConfigMap
metadata:
name: amazon-vpc-cni
namespace: kube-system
data:
enable-windows-ipam: "true"
  • Windows nodes will automatically be tainted
kubernetes_taints = [{
  key    = "WINDOWS"
  value  = "true"
  effect = "NO_SCHEDULE"
}]
  • Any pods that target Windows will need to have the following attributes set in their manifest
  nodeSelector:
    kubernetes.io/os: windows
    kubernetes.io/arch: amd64

https://docs.aws.amazon.com/eks/latest/userguide/windows-support.html

Variables​

Required Variables​

cluster_name (string) required

The name of the EKS cluster

desired_size (number) required

Initial desired number of worker nodes (external changes ignored)

max_size (number) required

Maximum number of worker nodes

min_size (number) required

Minimum number of worker nodes

subnet_ids (list(string)) required

A list of subnet IDs to launch resources in

Optional Variables​

after_cluster_joining_userdata (list(string)) optional

Additional bash commands to execute on each worker node after joining the EKS cluster (after executing the bootstrap.sh script). For more info, see https://kubedex.com/90-days-of-aws-eks-in-production


Default value: [ ]

ami_image_id (list(string)) optional

AMI to use, overriding other AMI specifications, but must match ami_type. Ignored if launch_template_id is supplied.


Default value: [ ]

ami_release_version (list(string)) optional

The EKS AMI "release version" to use. Defaults to the latest recommended version.
For Amazon Linux, it is the "Release version" from Amazon AMI Releases
For Bottlerocket, it is the release tag from Bottlerocket Releases without the "v" prefix.
For Windows, it is "AMI version" from AWS docs.
Note that unlike AMI names, release versions never include the "v" prefix.
Examples:
AL2: 1.29.3-20240531
Bottlerocket: 1.2.0 or 1.2.0-ccf1b754
Windows: 1.29-2024.04.09



Default value: [ ]

ami_type (string) optional

Type of Amazon Machine Image (AMI) associated with the EKS Node Group.
Defaults to AL2_x86_64. Valid values: AL2_x86_64, AL2_x86_64_GPU, AL2_ARM_64, CUSTOM, BOTTLEROCKET_ARM_64, BOTTLEROCKET_x86_64, BOTTLEROCKET_ARM_64_NVIDIA, BOTTLEROCKET_x86_64_NVIDIA, WINDOWS_CORE_2019_x86_64, WINDOWS_FULL_2019_x86_64, WINDOWS_CORE_2022_x86_64, WINDOWS_FULL_2022_x86_64, AL2023_x86_64_STANDARD, AL2023_ARM_64_STANDARD.



Default value: "AL2_x86_64"

associate_cluster_security_group (bool) optional

When true, associate the default cluster security group to the nodes. If disabled the EKS managed security group will not
be associated to the nodes and you will need to provide another security group that allows the nodes to communicate with
the EKS control plane. Be aware that if no associated_security_group_ids or ssh_access_security_group_ids are provided,
then the nodes will have no inbound or outbound rules.



Default value: true

associated_security_group_ids (list(string)) optional

A list of IDs of Security Groups to associate the node group with, in addition to the EKS' created security group.
These security groups will not be modified.



Default value: [ ]

before_cluster_joining_userdata (list(string)) optional

Additional bash commands to execute on each worker node before joining the EKS cluster (before executing the bootstrap.sh script). For more info, see https://kubedex.com/90-days-of-aws-eks-in-production


Default value: [ ]

block_device_map optional

Map of block device name specification, see launch_template.block-devices.



Type:

map(object({
no_device    = optional(bool, null)
virtual_name = optional(string, null)
ebs = optional(object({
  delete_on_termination = optional(bool, true)
  encrypted             = optional(bool, true)
  iops                  = optional(number, null)
  kms_key_id            = optional(string, null)
  snapshot_id           = optional(string, null)
  throughput            = optional(number, null)
  volume_size           = optional(number, 20)
  volume_type           = optional(string, "gp3")
}))
}))

Default value:

{
  "/dev/xvda": {
    "ebs": {}
  }
}
block_device_mappings (list(any)) optional

DEPRECATED: Use block_device_map instead.
List of block device mappings for the launch template.
Each list element is an object with a device_name key and
any keys supported by the ebs block of launch_template.



Default value: null

bootstrap_additional_options (list(string)) optional

Additional options to bootstrap.sh. DO NOT include --kubelet-additional-args, use kubelet_additional_options var instead. Not used with AL2023 AMI types.


Default value: [ ]

capacity_type (string) optional

Type of capacity associated with the EKS Node Group. Valid values: "ON_DEMAND", "SPOT", or null.
Terraform will only perform drift detection if a configuration value is provided.



Default value: null

cluster_autoscaler_enabled (bool) optional

OBSOLETE. Used to add support for the Kubernetes Cluster Autoscaler, but additional support is no longer needed.



Default value: null

cpu_options (list(any)) optional

Configuration for the cpu_options Configuration Block of the launch template.
Leave list empty for defaults. Pass list with single object with attributes matching the cpu_options block to configure it.
Note that this configures the launch template only. Some elements will be ignored by the Auto Scaling Group
that actually launches instances. Consult AWS documentation for details.



Default value: [ ]

create_before_destroy (bool) optional

If true (default), a new node group will be created before destroying the old one.
If false, the old node group will be destroyed first, causing downtime.
Changing this setting will always cause node group to be replaced.



Default value: true

detailed_monitoring_enabled (bool) optional

The launched EC2 instance will have detailed monitoring enabled. Defaults to false


Default value: false

ebs_optimized (bool) optional

Set false to disable EBS optimization


Default value: true

ec2_ssh_key_name (list(string)) optional

SSH key pair name to use to access the worker nodes


Default value: [ ]

enclave_enabled (bool) optional

Set to true to enable Nitro Enclaves on the instance.


Default value: false

force_update_version (bool) optional

When updating the Kubernetes version, force Pods to be removed even if PodDisruptionBudget or taint/toleration issues would otherwise prevent them from being removed (and cause the update to fail)


Default value: false

immediately_apply_lt_changes (bool) optional

When true, any change to the launch template will be applied immediately.
When false, the changes will only affect new nodes when they are launched.
When null (default) this input takes the value of create_before_destroy.
NOTE: Setting this to false does not guarantee that other changes,
such as ami_type, will not cause changes to be applied immediately.



Default value: null

instance_types (list(string)) optional

Instance types to use for this node group (up to 20). Defaults to ["t3.medium"].
Must be empty if the launch template configured by launch_template_id specifies an instance type.



Default value:

[
  "t3.medium"
]
kubelet_additional_options (list(string)) optional

Additional flags to pass to kubelet.
DO NOT include --node-labels or --node-taints,
use kubernetes_labels and kubernetes_taints to specify those."



Default value: [ ]

kubernetes_labels (map(string)) optional

Key-value mapping of Kubernetes labels. Only labels that are applied with the EKS API are managed by this argument.
Other Kubernetes labels applied to the EKS Node Group will not be managed.



Default value: { }

kubernetes_taints optional

List of key, value, effect objects representing Kubernetes taints.
effect must be one of NO_SCHEDULE, NO_EXECUTE, or PREFER_NO_SCHEDULE.
key and effect are required, value may be null.



Type:

list(object({
key    = string
value  = optional(string)
effect = string
}))

Default value: [ ]

kubernetes_version (list(string)) optional

Kubernetes version. Defaults to EKS Cluster Kubernetes version. Terraform will only perform drift detection if a configuration value is provided


Default value: [ ]

launch_template_id (list(string)) optional

The ID (not name) of a custom launch template to use for the EKS node group. If provided, it must specify the AMI image ID.


Default value: [ ]

launch_template_version (list(string)) optional

The version of the specified launch template to use. Defaults to latest version.


Default value: [ ]

metadata_http_endpoint_enabled (bool) optional

Set false to disable the Instance Metadata Service.


Default value: true

metadata_http_put_response_hop_limit (number) optional

The desired HTTP PUT response hop limit (between 1 and 64) for Instance Metadata Service requests.
The default is 2 to allows containerized workloads assuming the instance profile, but it's not really recomended. You should use OIDC service accounts instead.



Default value: 2

metadata_http_tokens_required (bool) optional

Set true to require IMDS session tokens, disabling Instance Metadata Service Version 1.


Default value: true

module_depends_on (any) optional

Can be any value desired. Module will wait for this value to be computed before creating node group.


Default value: null

node_group_terraform_timeouts optional

Configuration for the Terraform timeouts Configuration Block of the node group resource.
Leave list empty for defaults. Pass list with single object with attributes matching the timeouts block to configure it.
Leave attribute values null to preserve individual defaults while setting others.



Type:

list(object({
create = optional(string)
update = optional(string)
delete = optional(string)
}))

Default value: [ ]

node_role_arn (list(string)) optional

If provided, assign workers the given role, which this module will not modify


Default value: [ ]

node_role_cni_policy_enabled (bool) optional

When true, the AmazonEKS_CNI_Policy will be attached to the node IAM role.
This used to be required, but it is now recommended that this policy be
attached only to the aws-node Kubernetes service account. However, that
is difficult to do with Terraform, so this module defaults to the old pattern.



Default value: true

node_role_permissions_boundary (string) optional

If provided, all IAM roles will be created with this permissions boundary attached.


Default value: null

node_role_policy_arns (list(string)) optional

List of policy ARNs to attach to the worker role this module creates in addition to the default ones


Default value: [ ]

placement (list(any)) optional

Configuration for the placement Configuration Block of the launch template.
Leave list empty for defaults. Pass list with single object with attributes matching the placement block to configure it.
Note that this configures the launch template only. Some elements will be ignored by the Auto Scaling Group
that actually launches instances. Consult AWS documentation for details.



Default value: [ ]

random_pet_length (number) optional

In order to support "create before destroy" behavior, this module uses the random_pet
resource to generate a unique pet name for the node group, since the node group name
must be unique, meaning the new node group must have a different name than the old one.
This variable controls the length of the pet name, meaning the number of pet names
concatenated together. This module defaults to 1, but there are only 452 names available,
so users with large numbers of node groups may want to increase this value.



Default value: 1

replace_node_group_on_version_update (bool) optional

Force Node Group replacement when updating to a new Kubernetes version. If set to false (the default), the Node Groups will be updated in-place


Default value: false

resources_to_tag (list(string)) optional

List of auto-launched resource types to tag. Valid types are "instance", "volume", "elastic-gpu", "spot-instances-request", "network-interface".


Default value:

[
  "instance",
  "volume",
  "network-interface"
]
ssh_access_security_group_ids (list(string)) optional

Set of EC2 Security Group IDs to allow SSH access (port 22) to the worker nodes. If you specify ec2_ssh_key, but do not specify this configuration when you create an EKS Node Group, port 22 on the worker nodes is opened to the Internet (0.0.0.0/0)


Default value: [ ]

update_config (list(map(number))) optional

Configuration for the eks_node_group update_config Configuration Block.
Specify exactly one of max_unavailable (node count) or max_unavailable_percentage (percentage of nodes).



Default value: [ ]

userdata_override_base64 (list(string)) optional

Many features of this module rely on the bootstrap.sh provided with Amazon Linux, and this module
may generate "user data" that expects to find that script. If you want to use an AMI that is not
compatible with the userdata generated by this module, then use userdata_override_base64 to provide
your own (Base64 encoded) user data. Use "" to prevent any user data from being set.


Setting userdata_override_base64 disables kubernetes_taints, kubelet_additional_options,
before_cluster_joining_userdata, after_cluster_joining_userdata, and bootstrap_additional_options.



Default value: [ ]

Context Variables​

The following variables are defined in the context.tf file of this module and part of the terraform-null-label pattern.

additional_tag_map (map(string)) optional

Additional key-value pairs to add to each map in tags_as_list_of_maps. Not added to tags or id.
This is for some rare cases where resources want additional configuration of tags
and therefore take a list of maps with tag key, value, and additional configuration.


Required: No

Default value: { }

attributes (list(string)) optional

ID element. Additional attributes (e.g. workers or cluster) to add to id,
in the order they appear in the list. New attributes are appended to the
end of the list. The elements of the list are joined by the delimiter
and treated as a single ID element.


Required: No

Default value: [ ]

context (any) optional

Single object for setting entire context at once.
See description of individual variables for details.
Leave string and numeric variables as null to use default value.
Individual variable settings (non-null) override settings in context object,
except for attributes, tags, and additional_tag_map, which are merged.


Required: No

Default value:

{
  "additional_tag_map": {},
  "attributes": [],
  "delimiter": null,
  "descriptor_formats": {},
  "enabled": true,
  "environment": null,
  "id_length_limit": null,
  "label_key_case": null,
  "label_order": [],
  "label_value_case": null,
  "labels_as_tags": [
    "unset"
  ],
  "name": null,
  "namespace": null,
  "regex_replace_chars": null,
  "stage": null,
  "tags": {},
  "tenant": null
}
delimiter (string) optional

Delimiter to be used between ID elements.
Defaults to - (hyphen). Set to "" to use no delimiter at all.


Required: No

Default value: null

descriptor_formats (any) optional

Describe additional descriptors to be output in the descriptors output map.
Map of maps. Keys are names of descriptors. Values are maps of the form
\{<br/> format = string<br/> labels = list(string)<br/> \}
(Type is any so the map values can later be enhanced to provide additional options.)
format is a Terraform format string to be passed to the format() function.
labels is a list of labels, in order, to pass to format() function.
Label values will be normalized before being passed to format() so they will be
identical to how they appear in id.
Default is {} (descriptors output will be empty).


Required: No

Default value: { }

enabled (bool) optional

Set to false to prevent the module from creating any resources
Required: No

Default value: null

environment (string) optional

ID element. Usually used for region e.g. 'uw2', 'us-west-2', OR role 'prod', 'staging', 'dev', 'UAT'
Required: No

Default value: null

id_length_limit (number) optional

Limit id to this many characters (minimum 6).
Set to 0 for unlimited length.
Set to null for keep the existing setting, which defaults to 0.
Does not affect id_full.


Required: No

Default value: null

label_key_case (string) optional

Controls the letter case of the tags keys (label names) for tags generated by this module.
Does not affect keys of tags passed in via the tags input.
Possible values: lower, title, upper.
Default value: title.


Required: No

Default value: null

label_order (list(string)) optional

The order in which the labels (ID elements) appear in the id.
Defaults to ["namespace", "environment", "stage", "name", "attributes"].
You can omit any of the 6 labels ("tenant" is the 6th), but at least one must be present.


Required: No

Default value: null

label_value_case (string) optional

Controls the letter case of ID elements (labels) as included in id,
set as tag values, and output by this module individually.
Does not affect values of tags passed in via the tags input.
Possible values: lower, title, upper and none (no transformation).
Set this to title and set delimiter to "" to yield Pascal Case IDs.
Default value: lower.


Required: No

Default value: null

labels_as_tags (set(string)) optional

Set of labels (ID elements) to include as tags in the tags output.
Default is to include all labels.
Tags with empty values will not be included in the tags output.
Set to [] to suppress all generated tags.
Notes:
The value of the name tag, if included, will be the id, not the name.
Unlike other null-label inputs, the initial setting of labels_as_tags cannot be
changed in later chained modules. Attempts to change it will be silently ignored.


Required: No

Default value:

[
  "default"
]
name (string) optional

ID element. Usually the component or solution name, e.g. 'app' or 'jenkins'.
This is the only ID element not also included as a tag.
The "name" tag is set to the full id string. There is no tag with the value of the name input.


Required: No

Default value: null

namespace (string) optional

ID element. Usually an abbreviation of your organization name, e.g. 'eg' or 'cp', to help ensure generated IDs are globally unique
Required: No

Default value: null

regex_replace_chars (string) optional

Terraform regular expression (regex) string.
Characters matching the regex will be removed from the ID elements.
If not set, "/[^a-zA-Z0-9-]/" is used to remove all characters other than hyphens, letters and digits.


Required: No

Default value: null

stage (string) optional

ID element. Usually used to indicate role, e.g. 'prod', 'staging', 'source', 'build', 'test', 'deploy', 'release'
Required: No

Default value: null

tags (map(string)) optional

Additional tags (e.g. {'BusinessUnit': 'XYZ'}).
Neither the tag keys nor the tag values will be modified by this module.


Required: No

Default value: { }

tenant (string) optional

ID element (Rarely used, not included by default). A customer identifier, indicating who this instance of a resource is for
Required: No

Default value: null

Outputs​

WARNING_cluster_autoscaler_enabled

WARNING

eks_node_group_ami_id

The ID of the AMI used for the worker nodes, if specified

eks_node_group_arn

Amazon Resource Name (ARN) of the EKS Node Group

eks_node_group_cbd_pet_name

The pet name of this node group, if this module generated one

eks_node_group_id

EKS Cluster name and EKS Node Group name separated by a colon

eks_node_group_launch_template_id

The ID of the launch template used for this node group

eks_node_group_launch_template_name

The name of the launch template used for this node group

eks_node_group_remote_access_security_group_id

The ID of the security group generated to allow SSH access to the nodes, if this module generated one

eks_node_group_resources

List of objects containing information about underlying resources of the EKS Node Group

eks_node_group_role_arn

ARN of the worker nodes IAM role

eks_node_group_role_name

Name of the worker nodes IAM role

eks_node_group_status

Status of the EKS Node Group

eks_node_group_tags_all

A map of tags assigned to the resource, including those inherited from the provider default_tags configuration block.

Dependencies​

Requirements​

  • terraform, version: >= 1.3.0
  • aws, version: >= 5.8
  • random, version: >= 2.0

Providers​

  • aws, version: >= 5.8
  • random, version: >= 2.0

Modules​

NameVersionSourceDescription
label0.25.0cloudposse/label/nulln/a
ssh_access2.2.0cloudposse/security-group/awsn/a
this0.25.0cloudposse/label/nulln/a

Resources​

The following resources are used by this module:

Data Sources​

The following data sources are used by this module: