karpenter-node-pool

This component deploys Karpenter NodePools to an EKS cluster.

Karpenter is still in v0 and rapidly evolving. At this time, this component only supports a subset of the features available in Karpenter. Support could be added for additional features as needed.

Not supported:

• Elements of NodePool:
  • template.spec.kubelet
  • limits currently only supports cpu and memory. Other limits such as nvidia.com/gpu are not supported.
• Elements of NodeClass:
  • subnetSelectorTerms. This component only supports selecting all public or all private subnets of the referenced EKS cluster.
  • securityGroupSelectorTerms. This component only supports selecting the security group of the referenced EKS cluster.
  • amiSelectorTerms. Such terms override the amiFamily setting, which is the only AMI selection supported by this component.
  • instanceStorePolicy
  • userData
  • detailedMonitoring
  • associatePublicIPAddress

Usage

Stack Level: Regional

If provisioning more than one NodePool, it is best practice to create NodePools that are mutually exclusive or weighted.

components:
  terraform:
    eks/karpenter-node-pool:
      settings:
        spacelift:
          workspace_enabled: true
      vars:
        enabled: true
        eks_component_name: eks/cluster
        name: "karpenter-node-pool"
        # https://karpenter.sh/v0.36.0/docs/concepts/nodepools/
        node_pools:
          default:
            name: default
            # Whether to place EC2 instances launched by Karpenter into VPC private subnets. Set it to `false` to use public subnets
            private_subnets_enabled: true
            disruption:
              consolidation_policy: WhenUnderutilized
              consolidate_after: 1h
              max_instance_lifetime: 336h
              budgets:
                # This budget allows 0 disruptions during business hours (from 9am to 5pm) on weekdays
                - schedule: "0 9 * * mon-fri"
                  duration: 8h
                  nodes: "0"
            # The total cpu of the cluster. Maps to spec.limits.cpu in the Karpenter NodeClass
            total_cpu_limit: "100"
            # The total memory of the cluster. Maps to spec.limits.memory in the Karpenter NodeClass
            total_memory_limit: "1000Gi"
            # The weight of the node pool. See https://karpenter.sh/docs/concepts/scheduling/#weighted-nodepools
            weight: 50
            # Taints to apply to the nodes in the node pool. See https://karpenter.sh/docs/concepts/nodeclasses/#spectaints
            taints:
              - key: "node.kubernetes.io/unreachable"
                effect: "NoExecute"
                value: "true"
            # Taints to apply to the nodes in the node pool at startup. See https://karpenter.sh/docs/concepts/nodeclasses/#specstartuptaints
            startup_taints:
              - key: "node.kubernetes.io/unreachable"
                effect: "NoExecute"
                value: "true"
            # Metadata options for the node pool. See https://karpenter.sh/docs/concepts/nodeclasses/#specmetadataoptions
            metadata_options:
              httpEndpoint: "enabled" # allows the node to call the AWS metadata service
              httpProtocolIPv6: "disabled"
              httpPutResponseHopLimit: 2
              httpTokens: "required"
            # The AMI used by Karpenter provisioner when provisioning nodes. Based on the value set for amiFamily, Karpenter will automatically query for the appropriate EKS optimized AMI via AWS Systems Manager (SSM)
            # Bottlerocket, AL2, Ubuntu
            # https://karpenter.sh/v0.18.0/aws/provisioning/#amazon-machine-image-ami-family
            ami_family: AL2
            # Karpenter provisioner block device mappings.
            block_device_mappings:
              - deviceName: /dev/xvda
                ebs:
                  volumeSize: 200Gi
                  volumeType: gp3
                  encrypted: true
                  deleteOnTermination: true
            # Set acceptable (In) and unacceptable (Out) Kubernetes and Karpenter values for node provisioning based on
            # Well-Known Labels and cloud-specific settings. These can include instance types, zones, computer architecture,
            # and capacity type (such as AWS spot or on-demand).
            # See https://karpenter.sh/v0.18.0/provisioner/#specrequirements for more details
            requirements:
              - key: "karpenter.sh/capacity-type"
                operator: "In"
                values:
                  - "on-demand"
                  - "spot"
              - key: "node.kubernetes.io/instance-type"
                operator: "In"
                # See https://aws.amazon.com/ec2/instance-explorer/ and https://aws.amazon.com/ec2/instance-types/
                # Values limited by DenyEC2InstancesWithoutEncryptionInTransit service control policy
                # See https://github.com/cloudposse/terraform-aws-service-control-policies/blob/master/catalog/ec2-policies.yaml
                # Karpenter recommends allowing at least 20 instance types to ensure availability.
                values:
                  - "c5n.2xlarge"
                  - "c5n.xlarge"
                  - "c5n.large"
                  - "c6i.2xlarge"
                  - "c6i.xlarge"
                  - "c6i.large"
                  - "m5n.2xlarge"
                  - "m5n.xlarge"
                  - "m5n.large"
                  - "m5zn.2xlarge"
                  - "m5zn.xlarge"
                  - "m5zn.large"
                  - "m6i.2xlarge"
                  - "m6i.xlarge"
                  - "m6i.large"
                  - "r5n.2xlarge"
                  - "r5n.xlarge"
                  - "r5n.large"
                  - "r6i.2xlarge"
                  - "r6i.xlarge"
                  - "r6i.large"
              - key: "kubernetes.io/arch"
                operator: "In"
                values:
                  - "amd64"

Requirements

Name	Version
terraform	>= 1.3.0
aws	>= 4.9.0
helm	>= 2.0
kubernetes	>= 2.7.1, != 2.21.0

Providers

Name	Version
aws	>= 4.9.0
kubernetes	>= 2.7.1, != 2.21.0

Modules

Name	Source	Version
eks	cloudposse/stack-config/yaml//modules/remote-state	1.5.0
iam_roles	../../account-map/modules/iam-roles	n/a
this	cloudposse/label/null	0.25.0
vpc	cloudposse/stack-config/yaml//modules/remote-state	1.5.0

Resources

Name	Type
kubernetes_manifest.ec2_node_class	resource
kubernetes_manifest.node_pool	resource
aws_eks_cluster_auth.eks	data source

Inputs

Name	Description	Type	Default	Required
additional_tag_map	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.	map(string)	{}	no
attributes	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.	list(string)	[]	no
context	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.	any	{ "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 }	no
delimiter	Delimiter to be used between ID elements. Defaults to - (hyphen). Set to "" to use no delimiter at all.	string	null	no
descriptor_formats	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).	any	{}	no
eks_component_name	The name of the eks component	string	"eks/cluster"	no
enabled	Set to false to prevent the module from creating any resources	bool	null	no
environment	ID element. Usually used for region e.g. 'uw2', 'us-west-2', OR role 'prod', 'staging', 'dev', 'UAT'	string	null	no
helm_manifest_experiment_enabled	Enable storing of the rendered manifest for helm_release so the full diff of what is changing can been seen in the plan	bool	false	no
id_length_limit	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.	number	null	no
import_profile_name	AWS Profile name to use when importing a resource	string	null	no
import_role_arn	IAM Role ARN to use when importing a resource	string	null	no
kube_data_auth_enabled	If true, use an aws_eks_cluster_auth data source to authenticate to the EKS cluster. Disabled by kubeconfig_file_enabled or kube_exec_auth_enabled.	bool	false	no
kube_exec_auth_aws_profile	The AWS config profile for aws eks get-token to use	string	""	no
kube_exec_auth_aws_profile_enabled	If true, pass kube_exec_auth_aws_profile as the profile to aws eks get-token	bool	false	no
kube_exec_auth_enabled	If true, use the Kubernetes provider exec feature to execute aws eks get-token to authenticate to the EKS cluster. Disabled by kubeconfig_file_enabled, overrides kube_data_auth_enabled.	bool	true	no
kube_exec_auth_role_arn	The role ARN for aws eks get-token to use	string	""	no
kube_exec_auth_role_arn_enabled	If true, pass kube_exec_auth_role_arn as the role ARN to aws eks get-token	bool	true	no
kubeconfig_context	Context to choose from the Kubernetes config file. If supplied, kubeconfig_context_format will be ignored.	string	""	no
kubeconfig_context_format	A format string to use for creating the kubectl context name when kubeconfig_file_enabled is true and kubeconfig_context is not supplied. Must include a single %s which will be replaced with the cluster name.	string	""	no
kubeconfig_exec_auth_api_version	The Kubernetes API version of the credentials returned by the exec auth plugin	string	"client.authentication.k8s.io/v1beta1"	no
kubeconfig_file	The Kubernetes provider config_path setting to use when kubeconfig_file_enabled is true	string	""	no
kubeconfig_file_enabled	If true, configure the Kubernetes provider with kubeconfig_file and use that kubeconfig file for authenticating to the EKS cluster	bool	false	no
label_key_case	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.	string	null	no
label_order	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.	list(string)	null	no
label_value_case	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.	string	null	no
labels_as_tags	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.	set(string)	[ "default" ]	no
name	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.	string	null	no
namespace	ID element. Usually an abbreviation of your organization name, e.g. 'eg' or 'cp', to help ensure generated IDs are globally unique	string	null	no
node_pools	Configuration for node pools. See code for details.	map(object({ # The name of the Karpenter provisioner. The map key is used if this is not set. name = optional(string) # Whether to place EC2 instances launched by Karpenter into VPC private subnets. Set it to false to use public subnets. private_subnets_enabled = bool # The Disruption spec controls how Karpenter scales down the node group. # See the example (sadly not the specific spec.disruption documentation) at https://karpenter.sh/docs/concepts/nodepools/ for details disruption = optional(object({ # Describes which types of Nodes Karpenter should consider for consolidation. # If using 'WhenUnderutilized', Karpenter will consider all nodes for consolidation and attempt to remove or # replace Nodes when it discovers that the Node is underutilized and could be changed to reduce cost. # If using WhenEmpty, Karpenter will only consider nodes for consolidation that contain no workload pods. consolidation_policy = optional(string, "WhenUnderutilized") # The amount of time Karpenter should wait after discovering a consolidation decision (go duration string, s, m, or h). # This value can currently (v0.36.0) only be set when the consolidationPolicy is 'WhenEmpty'. # You can choose to disable consolidation entirely by setting the string value 'Never' here. # Earlier versions of Karpenter called this field ttl_seconds_after_empty. consolidate_after = optional(string) # The amount of time a Node can live on the cluster before being removed (go duration string, s, m, or h). # You can choose to disable expiration entirely by setting the string value 'Never' here. # This module sets a default of 336 hours (14 days), while the Karpenter default is 720 hours (30 days). # Note that Karpenter calls this field "expiresAfter", and earlier versions called it ttl_seconds_until_expired, # but we call it "max_instance_lifetime" to match the corresponding field in EC2 Auto Scaling Groups. max_instance_lifetime = optional(string, "336h") # Budgets control the the maximum number of NodeClaims owned by this NodePool that can be terminating at once. # See https://karpenter.sh/docs/concepts/disruption/#disruption-budgets for details. # A percentage is the percentage of the total number of active, ready nodes not being deleted, rounded up. # If there are multiple active budgets, Karpenter uses the most restrictive value. # If left undefined, this will default to one budget with a value of nodes: 10%. # Note that budgets do not prevent or limit involuntary terminations. # Example: # On Weekdays during business hours, don't do any deprovisioning. # budgets = { # schedule = "0 9 * * mon-fri" # duration = 8h # nodes = "0" # } budgets = optional(list(object({ # The schedule specifies when a budget begins being active, using extended cronjob syntax. # See https://kubernetes.io/docs/concepts/workloads/controllers/cron-jobs/#schedule-syntax for syntax details. # Timezones are not supported. This field is required if Duration is set. schedule = optional(string) # Duration determines how long a Budget is active after each Scheduled start. # If omitted, the budget is always active. This is required if Schedule is set. # Must be a whole number of minutes and hours, as cron does not work in seconds, # but since Go's duration.String() always adds a "0s" at the end, that is allowed. duration = optional(string) # The percentage or number of nodes that Karpenter can scale down during the budget. nodes = string })), []) }), {}) # Karpenter provisioner total CPU limit for all pods running on the EC2 instances launched by Karpenter total_cpu_limit = string # Karpenter provisioner total memory limit for all pods running on the EC2 instances launched by Karpenter total_memory_limit = string # Set a weight for this node pool. # See https://karpenter.sh/docs/concepts/scheduling/#weighted-nodepools weight = optional(number, 50) labels = optional(map(string)) annotations = optional(map(string)) # Karpenter provisioner taints configuration. See https://aws.github.io/aws-eks-best-practices/karpenter/#create-provisioners-that-are-mutually-exclusive for more details taints = optional(list(object({ key = string effect = string value = optional(string) }))) startup_taints = optional(list(object({ key = string effect = string value = optional(string) }))) # Karpenter node metadata options. See https://karpenter.sh/docs/concepts/nodeclasses/#specmetadataoptions for more details metadata_options = optional(object({ httpEndpoint = optional(string, "enabled") httpProtocolIPv6 = optional(string, "disabled") httpPutResponseHopLimit = optional(number, 2) # httpTokens can be either "required" or "optional" httpTokens = optional(string, "required") }), {}) # The AMI used by Karpenter provisioner when provisioning nodes. Based on the value set for amiFamily, Karpenter will automatically query for the appropriate EKS optimized AMI via AWS Systems Manager (SSM) ami_family = string # Karpenter nodes block device mappings. Controls the Elastic Block Storage volumes that Karpenter attaches to provisioned nodes. # Karpenter uses default block device mappings for the AMI Family specified. # For example, the Bottlerocket AMI Family defaults with two block device mappings, # and normally you only want to scale /dev/xvdb where Containers and there storage are stored. # Most other AMIs only have one device mapping at /dev/xvda. # See https://karpenter.sh/docs/concepts/nodeclasses/#specblockdevicemappings for more details block_device_mappings = list(object({ deviceName = string ebs = optional(object({ volumeSize = string volumeType = string deleteOnTermination = optional(bool, true) encrypted = optional(bool, true) iops = optional(number) kmsKeyID = optional(string, "alias/aws/ebs") snapshotID = optional(string) throughput = optional(number) })) })) # Set acceptable (In) and unacceptable (Out) Kubernetes and Karpenter values for node provisioning based on Well-Known Labels and cloud-specific settings. These can include instance types, zones, computer architecture, and capacity type (such as AWS spot or on-demand). See https://karpenter.sh/v0.18.0/provisioner/#specrequirements for more details requirements = list(object({ key = string operator = string # Operators like "Exists" and "DoesNotExist" do not require a value values = optional(list(string)) })) # Any values for spec.template.spec.kubelet allowed by Karpenter. # Not fully specified, because they are subject to change. # See: # https://karpenter.sh/docs/concepts/nodepools/#spectemplatespeckubelet # https://kubernetes.io/docs/reference/config-api/kubelet-config.v1beta1/ kubelet = optional(any, {}) }))	n/a	yes
regex_replace_chars	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.	string	null	no
region	AWS Region	string	n/a	yes
stage	ID element. Usually used to indicate role, e.g. 'prod', 'staging', 'source', 'build', 'test', 'deploy', 'release'	string	null	no
tags	Additional tags (e.g. {'BusinessUnit': 'XYZ'}). Neither the tag keys nor the tag values will be modified by this module.	map(string)	{}	no
tenant	ID element _(Rarely used, not included by default)_. A customer identifier, indicating who this instance of a resource is for	string	null	no

Outputs

Name	Description
ec2_node_classes	Deployed Karpenter EC2NodeClass
node_pools	Deployed Karpenter NodePool

References

• https://karpenter.sh
• https://aws.github.io/aws-eks-best-practices/karpenter
• https://karpenter.sh/docs/concepts/nodepools
• https://aws.amazon.com/blogs/aws/introducing-karpenter-an-open-source-high-performance-kubernetes-cluster-autoscaler
• https://github.com/aws/karpenter
• https://ec2spotworkshops.com/karpenter.html
• https://www.eksworkshop.com/docs/autoscaling/compute/karpenter/

Changelog

Release 1.470.0

Components PR #1076

• Allow specifying elements of spec.template.spec.kubelet
• Make taint values optional

The var.node_pools map now includes a kubelet field that allows specifying elements of spec.template.spec.kubelet. This is useful for configuring the kubelet to use custom settings, such as reserving resources for system daemons.

For more information, see:

• Karpenter documentation
• Kubernetes documentation

The value fields of the taints and startup_taints lists in the var.node_pools map are now optional. This is in alignment with the Kubernetes API, where key and effect are required, but the value field is optional.