Karpenter – Concepts

V0.32: NodePools

Mon, 01 Jan 0001 00:00:00 +0000

When you first installed Karpenter, you set up a default NodePool. The NodePool sets constraints on the nodes that can be created by Karpenter and the pods that can run on those nodes. The NodePool can be set to do things like:

Define taints to limit the pods that can run on nodes Karpenter creates
Define any startup taints to inform Karpenter that it should taint the node initially, but that the taint is temporary.
Limit node creation to certain zones, instance types, and computer architectures
Set defaults for node expiration

You can change your NodePool or add other NodePools to Karpenter. Here are things you should know about NodePools:

Karpenter won’t do anything if there is not at least one NodePool configured.
Each NodePool that is configured is looped through by Karpenter.
If Karpenter encounters a taint in the NodePool that is not tolerated by a Pod, Karpenter won’t use that NodePool to provision the pod.
If Karpenter encounters a startup taint in the NodePool it will be applied to nodes that are provisioned, but pods do not need to tolerate the taint. Karpenter assumes that the taint is temporary and some other system will remove the taint.
It is recommended to create NodePools that are mutually exclusive. So no Pod should match multiple NodePools. If multiple NodePools are matched, Karpenter will use the NodePool with the highest weight.

For some example NodePool configurations, see the examples in the Karpenter GitHub repository.

apiVersion: karpenter.sh/v1beta1
kind: NodePool
metadata:
 name: default
spec:
 # Template section that describes how to template out NodeClaim resources that Karpenter will provision
 # Karpenter will consider this template to be the minimum requirements needed to provision a Node using this NodePool
 # It will overlay this NodePool with Pods that need to schedule to further constrain the NodeClaims
 # Karpenter will provision to launch new Nodes for the cluster
 template:
 metadata:
 # Labels are arbitrary key-values that are applied to all nodes
 labels:
 billing-team: my-team

 # Annotations are arbitrary key-values that are applied to all nodes
 annotations:
 example.com/owner: "my-team"
 spec:
 # References the Cloud Provider's NodeClass resource, see your cloud provider specific documentation
 nodeClassRef:
 apiVersion: karpenter.k8s.aws/v1beta1
 kind: EC2NodeClass
 name: default

 # Provisioned nodes will have these taints
 # Taints may prevent pods from scheduling if they are not tolerated by the pod.
 taints:
 - key: example.com/special-taint
 effect: NoSchedule

 # Provisioned nodes will have these taints, but pods do not need to tolerate these taints to be provisioned by this
 # NodePool. These taints are expected to be temporary and some other entity (e.g. a DaemonSet) is responsible for
 # removing the taint after it has finished initializing the node.
 startupTaints:
 - key: example.com/another-taint
 effect: NoSchedule

 # Requirements that constrain the parameters of provisioned nodes.
 # These requirements are combined with pod.spec.topologySpreadConstraints, pod.spec.affinity.nodeAffinity, pod.spec.affinity.podAffinity, and pod.spec.nodeSelector rules.
 # Operators { In, NotIn, Exists, DoesNotExist, Gt, and Lt } are supported.
 # https://kubernetes.io/docs/concepts/scheduling-eviction/assign-pod-node/#operators
 requirements:
 - key: "karpenter.k8s.aws/instance-category"
 operator: In
 values: ["c", "m", "r"]
 - key: "karpenter.k8s.aws/instance-cpu"
 operator: In
 values: ["4", "8", "16", "32"]
 - key: "karpenter.k8s.aws/instance-hypervisor"
 operator: In
 values: ["nitro"]
 - key: "karpenter.k8s.aws/instance-generation"
 operator: Gt
 values: ["2"]
 - key: "topology.kubernetes.io/zone"
 operator: In
 values: ["us-west-2a", "us-west-2b"]
 - key: "kubernetes.io/arch"
 operator: In
 values: ["arm64", "amd64"]
 - key: "karpenter.sh/capacity-type"
 operator: In
 values: ["spot", "on-demand"]

 # Karpenter provides the ability to specify a few additional Kubelet args.
 # These are all optional and provide support for additional customization and use cases.
 kubelet:
 clusterDNS: ["10.0.1.100"]
 systemReserved:
 cpu: 100m
 memory: 100Mi
 ephemeral-storage: 1Gi
 kubeReserved:
 cpu: 200m
 memory: 100Mi
 ephemeral-storage: 3Gi
 evictionHard:
 memory.available: 5%
 nodefs.available: 10%
 nodefs.inodesFree: 10%
 evictionSoft:
 memory.available: 500Mi
 nodefs.available: 15%
 nodefs.inodesFree: 15%
 evictionSoftGracePeriod:
 memory.available: 1m
 nodefs.available: 1m30s
 nodefs.inodesFree: 2m
 evictionMaxPodGracePeriod: 60
 imageGCHighThresholdPercent: 85
 imageGCLowThresholdPercent: 80
 cpuCFSQuota: true
 podsPerCore: 2
 maxPods: 20

 # Disruption section which describes the ways in which Karpenter can disrupt and replace Nodes
 # Configuration in this section constrains how aggressive Karpenter can be with performing operations
 # like rolling Nodes due to them hitting their maximum lifetime (expiry) or scaling down nodes to reduce cluster cost
 disruption:
 # 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
 consolidationPolicy: WhenUnderutilized | WhenEmpty

 # The amount of time Karpenter should wait after discovering a consolidation decision
 # This value can currently only be set when the consolidationPolicy is 'WhenEmpty'
 # You can choose to disable consolidation entirely by setting the string value 'Never' here
 consolidateAfter: 30s

 # The amount of time a Node can live on the cluster before being removed
 # Avoiding long-running Nodes helps to reduce security vulnerabilities as well as to reduce the chance of issues that can plague Nodes with long uptimes such as file fragmentation or memory leaks from system processes
 # You can choose to disable expiration entirely by setting the string value 'Never' here
 expireAfter: 720h

 # Resource limits constrain the total size of the pool.
 # Limits prevent Karpenter from creating new instances once the limit is exceeded.
 limits:
 cpu: "1000"
 memory: 1000Gi

 # Priority given to the NodePool when the scheduler considers which NodePool
 # to select. Higher weights indicate higher priority when comparing NodePools.
 # Specifying no weight is equivalent to specifying a weight of 0.
 weight: 10

spec.template.spec.requirements

Kubernetes defines the following Well-Known Labels, and cloud providers (e.g., AWS) implement them. They are defined at the “spec.requirements” section of the NodePool API.

In addition to the well-known labels from Kubernetes, Karpenter supports AWS-specific labels for more advanced scheduling. See the full list here.

These well-known labels may be specified at the NodePool level, or in a workload definition (e.g., nodeSelector on a pod.spec). Nodes are chosen using both the NodePool’s and pod’s requirements. If there is no overlap, nodes will not be launched. In other words, a pod’s requirements must be within the NodePool’s requirements. If a requirement is not defined for a well known label, any value available to the cloud provider may be chosen.

For example, an instance type may be specified using a nodeSelector in a pod spec. If the instance type requested is not included in the NodePool list and the NodePool has instance type requirements, Karpenter will not create a node or schedule the pod.

Instance Types

key: node.kubernetes.io/instance-type
key: karpenter.k8s.aws/instance-family
key: karpenter.k8s.aws/instance-category
key: karpenter.k8s.aws/instance-generation

Generally, instance types should be a list and not a single value. Leaving these requirements undefined is recommended, as it maximizes choices for efficiently placing pods.

Review AWS instance types. Most instance types are supported with the exclusion of non-HVM.

Availability Zones

key: topology.kubernetes.io/zone
value example: us-east-1c
value list: aws ec2 describe-availability-zones --region <region-name>

Karpenter can be configured to create nodes in a particular zone. Note that the Availability Zone us-east-1a for your AWS account might not have the same location as us-east-1a for another AWS account.

Learn more about Availability Zone IDs.

Architecture

key: kubernetes.io/arch
values
- amd64
- arm64

Karpenter supports amd64 nodes, and arm64 nodes.

Operating System

key: kubernetes.io/os
values
- linux
- windows

Karpenter supports linux and windows operating systems.

Capacity Type

key: karpenter.sh/capacity-type
values
- spot
- on-demand

Karpenter supports specifying capacity type, which is analogous to EC2 purchase options.

Karpenter prioritizes Spot offerings if the NodePool allows Spot and on-demand instances (note that in this scenario any Spot instances priced higher than the cheapest on-demand instance will be temporarily removed from consideration). If the provider API (e.g. EC2 Fleet’s API) indicates Spot capacity is unavailable, Karpenter caches that result across all attempts to provision EC2 capacity for that instance type and zone for the next 3 minutes. If there are no other possible offerings available for Spot, Karpenter will attempt to provision on-demand instances, generally within milliseconds.

Karpenter also allows karpenter.sh/capacity-type to be used as a topology key for enforcing topology-spread.

Karpenter allows you to be extremely flexible with your NodePools by only constraining your instance types in ways that are absolutely necessary for your cluster. By default, Karpenter will enforce that you specify the spec.template.spec.requirements field, but will not enforce that you specify any requirements within the field. If you choose to specify requirements: [], this means that you will completely flexible to all instance types that your cloud provider supports.

Though Karpenter doesn’t enforce these defaults, for most use-cases, we recommend that you specify some requirements to avoid odd behavior or exotic instance types. Below, is a high-level recommendation for requirements that should fit the majority of use-cases for generic workloads

spec:
 template:
 spec:
 requirements:
 - key: kubernetes.io/arch
 operator: In
 values: ["amd64"]
 - key: kubernetes.io/os
 operator: In
 values: ["linux"]
 - key: karpenter.sh/capacity-type
 operator: In
 values: ["on-demand"]
 - key: karpenter.k8s.aws/instance-category
 operator: In
 values: ["c", "m", "r"]
 - key: karpenter.k8s.aws/instance-generation
 operator: Gt
 values: ["2"]

Note

There is currently a limit of 30 on the total number of requirements on both the NodePool and the NodeClaim. It’s important to note that spec.template.metadata.labels are also propagated as requirements on the NodeClaim when it’s created, meaning that you can’t have more than 30 requirements and labels combined set on your NodePool.

spec.template.spec.nodeClassRef

This field points to the Cloud Provider NodeClass resource. Learn more about EC2NodeClasses.

spec.template.spec.kubelet

Karpenter provides the ability to specify a few additional Kubelet args. These are all optional and provide support for additional customization and use cases. Adjust these only if you know you need to do so. For more details on kubelet configuration arguments, see the KubeletConfiguration API specification docs. The implemented fields are a subset of the full list of upstream kubelet configuration arguments. Please cut an issue if you’d like to see another field implemented.

kubelet:
 clusterDNS: ["10.0.1.100"]
 systemReserved:
 cpu: 100m
 memory: 100Mi
 ephemeral-storage: 1Gi
 kubeReserved:
 cpu: 200m
 memory: 100Mi
 ephemeral-storage: 3Gi
 evictionHard:
 memory.available: 5%
 nodefs.available: 10%
 nodefs.inodesFree: 10%
 evictionSoft:
 memory.available: 500Mi
 nodefs.available: 15%
 nodefs.inodesFree: 15%
 evictionSoftGracePeriod:
 memory.available: 1m
 nodefs.available: 1m30s
 nodefs.inodesFree: 2m
 evictionMaxPodGracePeriod: 60
 imageGCHighThresholdPercent: 85
 imageGCLowThresholdPercent: 80
 cpuCFSQuota: true
 podsPerCore: 2
 maxPods: 20

Reserved Resources

Karpenter will automatically configure the system and kube reserved resource requests on the fly on your behalf. These requests are used to configure your node and to make scheduling decisions for your pods. If you have specific requirements or know that you will have additional capacity requirements, you can optionally override the --system-reserved configuration defaults with the .spec.template.spec.kubelet.systemReserved values and the --kube-reserved configuration defaults with the .spec.template.spec.kubelet.kubeReserved values.

Note

Karpenter considers these reserved resources when computing the allocatable ephemeral storage on a given instance type. If kubeReserved is not specified, Karpenter will compute the default reserved CPU and memory resources for the purpose of ephemeral storage computation. These defaults are based on the defaults on Karpenter’s supported AMI families, which are not the same as the kubelet defaults. You should be aware of the CPU and memory default calculation when using Custom AMI Families. If they don’t align, there may be a difference in Karpenter’s computed allocatable ephemeral storage and the actually ephemeral storage available on the node.

Eviction Thresholds

The kubelet supports eviction thresholds by default. When enough memory or file system pressure is exerted on the node, the kubelet will begin to evict pods to ensure that system daemons and other system processes can continue to run in a healthy manner.

Kubelet has the notion of hard evictions and soft evictions. In hard evictions, pods are evicted as soon as a threshold is met, with no grace period to terminate. Soft evictions, on the other hand, provide an opportunity for pods to be terminated gracefully. They do so by sending a termination signal to pods that are planning to be evicted and allowing those pods to terminate up to their grace period.

Karpenter supports hard evictions through the .spec.template.spec.kubelet.evictionHard field and soft evictions through the .spec.template.spec.kubelet.evictionSoft field. evictionHard and evictionSoft are configured by listing signal names with either percentage values or resource values.

kubelet:
 evictionHard:
 memory.available: 500Mi
 nodefs.available: 10%
 nodefs.inodesFree: 10%
 imagefs.available: 5%
 imagefs.inodesFree: 5%
 pid.available: 7%
 evictionSoft:
 memory.available: 1Gi
 nodefs.available: 15%
 nodefs.inodesFree: 15%
 imagefs.available: 10%
 imagefs.inodesFree: 10%
 pid.available: 10%

Supported Eviction Signals

Eviction Signal	Description
memory.available	memory.available := node.status.capacity[memory] - node.stats.memory.workingSet
nodefs.available	nodefs.available := node.stats.fs.available
nodefs.inodesFree	nodefs.inodesFree := node.stats.fs.inodesFree
imagefs.available	imagefs.available := node.stats.runtime.imagefs.available
imagefs.inodesFree	imagefs.inodesFree := node.stats.runtime.imagefs.inodesFree
pid.available	pid.available := node.stats.rlimit.maxpid - node.stats.rlimit.curproc

For more information on eviction thresholds, view the Node-pressure Eviction section of the official Kubernetes docs.

Soft Eviction Grace Periods

Soft eviction pairs an eviction threshold with a specified grace period. With soft eviction thresholds, the kubelet will only begin evicting pods when the node exceeds its soft eviction threshold over the entire duration of its grace period. For example, if you specify evictionSoft[memory.available] of 500Mi and a evictionSoftGracePeriod[memory.available] of 1m30, the node must have less than 500Mi of available memory over a minute and a half in order for the kubelet to begin evicting pods.

Optionally, you can specify an evictionMaxPodGracePeriod which defines the administrator-specified maximum pod termination grace period to use during soft eviction. If a namespace-owner had specified a pod terminationGracePeriodInSeconds on pods in their namespace, the minimum of evictionPodGracePeriod and terminationGracePeriodInSeconds would be used.

kubelet:
 evictionSoftGracePeriod:
 memory.available: 1m
 nodefs.available: 1m30s
 nodefs.inodesFree: 2m
 imagefs.available: 1m30s
 imagefs.inodesFree: 2m
 pid.available: 2m
 evictionMaxPodGracePeriod: 60

Pod Density

By default, the number of pods on a node is limited by both the number of networking interfaces (ENIs) that may be attached to an instance type and the number of IP addresses that can be assigned to each ENI. See IP addresses per network interface per instance type for a more detailed information on these instance types’ limits.

Note

By default, the VPC CNI allocates IPs for a node and pods from the same subnet. With VPC CNI Custom Networking, the pods will receive IP addresses from another subnet dedicated to pod IPs. This approach makes it easier to manage IP addresses and allows for separate Network Access Control Lists (NACLs) applied to your pods. VPC CNI Custom Networking reduces the pod density of a node since one of the ENI attachments will be used for the node and cannot share the allocated IPs on the interface to pods. Karpenter supports VPC CNI Custom Networking and similar CNI setups where the primary node interface is separated from the pods interfaces through a global setting within the karpenter-global-settings configmap: aws.reservedENIs. In the common case, aws.reservedENIs should be set to "1" if using Custom Networking.

Windows Support Notice

It’s currently not possible to specify custom networking with Windows nodes.

Max Pods

For small instances that require an increased pod density or large instances that require a reduced pod density, you can override this default value with .spec.template.spec.kubelet.maxPods. This value will be used during Karpenter pod scheduling and passed through to --max-pods on kubelet startup.

Note

When using small instance types, it may be necessary to enable prefix assignment mode in the AWS VPC CNI plugin to support a higher pod density per node. Prefix assignment mode was introduced in AWS VPC CNI v1.9 and allows ENIs to manage a broader set of IP addresses. Much higher pod densities are supported as a result.

Windows Support Notice

Presently, Windows worker nodes do not support using more than one ENI. As a consequence, the number of IP addresses, and subsequently, the number of pods that a Windows worker node can support is limited by the number of IPv4 addresses available on the primary ENI. Currently, Karpenter will only consider individual secondary IP addresses when calculating the pod density limit.

Pods Per Core

An alternative way to dynamically set the maximum density of pods on a node is to use the .spec.template.spec.kubelet.podsPerCore value. Karpenter will calculate the pod density during scheduling by multiplying this value by the number of logical cores (vCPUs) on an instance type. This value will also be passed through to the --pods-per-core value on kubelet startup to configure the number of allocatable pods the kubelet can assign to the node instance.

The value generated from podsPerCore cannot exceed maxPods, meaning, if both are set, the minimum of the podsPerCore dynamic pod density and the static maxPods value will be used for scheduling.

Note

maxPods may not be set in the kubelet of a NodePool, but may still be restricted by the ENI_LIMITED_POD_DENSITY value. You may want to ensure that the podsPerCore value that will be used for instance families associated with the NodePool will not cause unexpected behavior by exceeding the maxPods value.

Pods Per Core on Bottlerocket

Bottlerocket AMIFamily currently does not support podsPerCore configuration. If a NodePool contains a provider or providerRef to a node template that will launch a Bottlerocket instance, the podsPerCore value will be ignored for scheduling and for configuring the kubelet.

spec.disruption

You can configure Karpenter to disrupt Nodes through your NodePool in multiple ways. You can use spec.disruption.consolidationPolicy, spec.disruption.consolidateAfter or spec.disruption.expireAfter. Read Disruption for more.

spec.limits

The NodePool spec includes a limits section (spec.limits), which constrains the maximum amount of resources that the NodePool can consume.

If the NodePool.spec.limits section is unspecified, it means that there is no default limitation on resource allocation. In this case, the maximum resource consumption is governed by the quotas set by your cloud provider. If a limit has been exceeded, nodes provisioning is prevented until some nodes have been terminated.

apiVersion: karpenter.sh/v1beta1
kind: NodePool
metadata:
 name: default
spec:
 template:
 spec:
 requirements:
 - key: karpenter.sh/capacity-type
 operator: In
 values: ["spot"]
 limits:
 cpu: 1000
 memory: 1000Gi
 nvidia.com/gpu: 2

Note

Karpenter provisioning is highly parallel. Because of this, limit checking is eventually consistent, which can result in overrun during rapid scale outs.

CPU limits are described with a DecimalSI value. Note that the Kubernetes API will coerce this into a string, so we recommend against using integers to avoid GitOps skew.

Memory limits are described with a BinarySI value, such as 1000Gi.

You can view the current consumption of cpu and memory on your cluster by running:

kubectl get nodepool -o=jsonpath='{.items[0].status}'

Review the Kubernetes core API (k8s.io/api/core/v1) for more information on resources.

spec.weight

Karpenter allows you to describe NodePool preferences through a weight mechanism similar to how weight is described with pod and node affinities.

For more information on weighting NodePools, see the Weighted NodePools section in the scheduling docs.

Examples

Isolating Expensive Hardware

A NodePool can be set up to only provision nodes on particular processor types. The following example sets a taint that only allows pods with tolerations for Nvidia GPUs to be scheduled:

apiVersion: karpenter.sh/v1beta1
kind: NodePool
metadata:
 name: gpu
spec:
 disruption:
 consolidationPolicy: WhenUnderutilized
 template:
 spec:
 requirements:
 - key: node.kubernetes.io/instance-type
 operator: In
 values: ["p3.8xlarge", "p3.16xlarge"]
 taints:
 - key: nvidia.com/gpu
 value: "true"
 effect: NoSchedule

In order for a pod to run on a node defined in this NodePool, it must tolerate nvidia.com/gpu in its pod spec.

Cilium Startup Taint

Per the Cilium docs, it’s recommended to place a taint of node.cilium.io/agent-not-ready=true:NoExecute on nodes to allow Cilium to configure networking prior to other pods starting. This can be accomplished via the use of Karpenter startupTaints. These taints are placed on the node, but pods aren’t required to tolerate these taints to be considered for provisioning.

apiVersion: karpenter.sh/v1beta1
kind: NodePool
metadata:
 name: cilium-startup
spec:
 disruption:
 consolidationPolicy: WhenUnderutilized
 template:
 spec:
 startupTaints:
 - key: node.cilium.io/agent-not-ready
 value: "true"
 effect: NoExecute

V0.32: NodeClasses

Mon, 01 Jan 0001 00:00:00 +0000

Node Classes enable configuration of AWS specific settings. Each NodePool must reference an EC2NodeClass using spec.template.spec.nodeClassRef. Multiple NodePools may point to the same EC2NodeClass.

apiVersion: karpenter.sh/v1beta1
kind: NodePool
metadata:
 name: default
spec:
 template:
 spec:
 nodeClassRef:
 apiVersion: karpenter.k8s.aws/v1beta1
 kind: EC2NodeClass
 name: default
---
apiVersion: karpenter.k8s.aws/v1beta1
kind: EC2NodeClass
metadata:
 name: default
spec:
 # Required, resolves a default ami and userdata
 amiFamily: AL2

 # Required, discovers subnets to attach to instances
 # Each term in the array of subnetSelectorTerms is ORed together
 # Within a single term, all conditions are ANDed
 subnetSelectorTerms:
 # Select on any subnet that has the "karpenter.sh/discovery: ${CLUSTER_NAME}"
 # AND the "environment: test" tag OR any subnet with ID "subnet-09fa4a0a8f233a921"
 - tags:
 karpenter.sh/discovery: "${CLUSTER_NAME}"
 environment: test
 - id: subnet-09fa4a0a8f233a921

 # Required, discovers security groups to attach to instances
 # Each term in the array of securityGroupSelectorTerms is ORed together
 # Within a single term, all conditions are ANDed
 securityGroupSelectorTerms:
 # Select on any security group that has both the "karpenter.sh/discovery: ${CLUSTER_NAME}" tag
 # AND the "environment: test" tag OR any security group with the "my-security-group" name
 # OR any security group with ID "sg-063d7acfb4b06c82c"
 - tags:
 karpenter.sh/discovery: "${CLUSTER_NAME}"
 environment: test
 - name: my-security-group
 - id: sg-063d7acfb4b06c82c

 # Optional, IAM role to use for the node identity.
 # The "role" field is immutable after EC2NodeClass creation. This may change in the
 # future, but this restriction is currently in place today to ensure that Karpenter
 # avoids leaking managed instance profiles in your account.
 # Must specify one of "role" or "instanceProfile" for Karpenter to launch nodes
 role: "KarpenterNodeRole-${CLUSTER_NAME}"

 # Optional, IAM instance profile to use for the node identity.
 # Must specify one of "role" or "instanceProfile" for Karpenter to launch nodes
 instanceProfile: "KarpenterNodeInstanceProfile-${CLUSTER_NAME}"

 # Optional, discovers amis to override the amiFamily's default amis
 # Each term in the array of amiSelectorTerms is ORed together
 # Within a single term, all conditions are ANDed
 amiSelectorTerms:
 # Select on any AMI that has both the "karpenter.sh/discovery: ${CLUSTER_NAME}" tag
 # AND the "environment: test" tag OR any AMI with the "my-ami" name
 # OR any AMI with ID "ami-123"
 - tags:
 karpenter.sh/discovery: "${CLUSTER_NAME}"
 environment: test
 - name: my-ami
 - id: ami-123

 # Optional, overrides autogenerated userdata with a merge semantic
 userData: |
 echo "Hello world"

 # Optional, propagates tags to underlying EC2 resources
 tags:
 team: team-a
 app: team-a-app

 # Optional, configures IMDS for the instance
 metadataOptions:
 httpEndpoint: enabled
 httpProtocolIPv6: disabled
 httpPutResponseHopLimit: 2
 httpTokens: required

 # Optional, configures storage devices for the instance
 blockDeviceMappings:
 - deviceName: /dev/xvda
 ebs:
 volumeSize: 100Gi
 volumeType: gp3
 iops: 10000
 encrypted: true
 kmsKeyID: "1234abcd-12ab-34cd-56ef-1234567890ab"
 deleteOnTermination: true
 throughput: 125
 snapshotID: snap-0123456789

 # Optional, configures detailed monitoring for the instance
 detailedMonitoring: true
status:
 # Resolved subnets
 subnets:
 - id: subnet-0a462d98193ff9fac
 zone: us-east-2b
 - id: subnet-0322dfafd76a609b6
 zone: us-east-2c
 - id: subnet-0727ef01daf4ac9fe
 zone: us-east-2b
 - id: subnet-00c99aeafe2a70304
 zone: us-east-2a
 - id: subnet-023b232fd5eb0028e
 zone: us-east-2c
 - id: subnet-03941e7ad6afeaa72
 zone: us-east-2a

 # Resolved security groups
 securityGroups:
 - id: sg-041513b454818610b
 name: ClusterSharedNodeSecurityGroup
 - id: sg-0286715698b894bca
 name: ControlPlaneSecurityGroup-1AQ073TSAAPW

 # Resolved AMIs
 amis:
 - id: ami-01234567890123456
 name: custom-ami-amd64
 requirements:
 - key: kubernetes.io/arch
 operator: In
 values:
 - amd64
 - id: ami-01234567890123456
 name: custom-ami-arm64
 requirements:
 - key: kubernetes.io/arch
 operator: In
 values:
 - arm64

 # Generated instance profile name from "role"
 instanceProfile: "${CLUSTER_NAME}-0123456778901234567789"

Refer to the NodePool docs for settings applicable to all providers. To explore various EC2NodeClass configurations, refer to the examples provided in the Karpenter Github repository.

spec.amiFamily

AMIFamily is a required field, dictating both the default bootstrapping logic for nodes provisioned through this EC2NodeClass but also selecting a group of recommended, latest AMIs by default. Currently, Karpenter supports amiFamily values AL2, Bottlerocket, Ubuntu, Windows2019, Windows2022 and Custom. GPUs are only supported by default with AL2 and Bottlerocket. The AL2 amiFamily does not support ARM64 GPU instance types unless you specify custom amiSelectorTerms. Default bootstrapping logic is shown below for each of the supported families.

AL2

MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="//"

--//
Content-Type: text/x-shellscript; charset="us-ascii"

#!/bin/bash -xe
exec > >(tee /var/log/user-data.log|logger -t user-data -s 2>/dev/console) 2>&1
/etc/eks/bootstrap.sh 'test-cluster' --apiserver-endpoint 'https://test-cluster' --b64-cluster-ca 'ca-bundle' \
--dns-cluster-ip '10.100.0.10' \
--use-max-pods false \
--container-runtime containerd \
--kubelet-extra-args '--node-labels=karpenter.sh/capacity-type=on-demand,karpenter.sh/nodepool=test --max-pods=110'
--//--

Bottlerocket

[settings]
[settings.kubernetes]
api-server = 'https://test-cluster'
cluster-certificate = 'ca-bundle'
cluster-name = 'test-cluster'
cluster-dns-ip = '10.100.0.10'
max-pods = 110

[settings.kubernetes.node-labels]
'karpenter.sh/capacity-type' = 'on-demand'
'karpenter.sh/nodepool' = 'test'

Ubuntu

MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="//"

--//
Content-Type: text/x-shellscript; charset="us-ascii"

#!/bin/bash -xe
exec > >(tee /var/log/user-data.log|logger -t user-data -s 2>/dev/console) 2>&1
/etc/eks/bootstrap.sh 'test-cluster' --apiserver-endpoint 'https://test-cluster' --b64-cluster-ca 'ca-bundle' \
--dns-cluster-ip '10.100.0.10' \
--use-max-pods false \
--kubelet-extra-args '--node-labels="karpenter.sh/capacity-type=on-demand,karpenter.sh/nodepool=test" --max-pods=110'
--//--

Windows2019

<powershell>
[string]$EKSBootstrapScriptFile = "$env:ProgramFiles\Amazon\EKS\Start-EKSBootstrap.ps1"
& $EKSBootstrapScriptFile -EKSClusterName 'test-cluster' -APIServerEndpoint 'https://test-cluster' -Base64ClusterCA 'ca-bundle' -KubeletExtraArgs '--node-labels="karpenter.sh/capacity-type=on-demand,karpenter.sh/nodepool=test" --max-pods=110' -DNSClusterIP '10.100.0.10'
</powershell>

Windows2022

<powershell>
[string]$EKSBootstrapScriptFile = "$env:ProgramFiles\Amazon\EKS\Start-EKSBootstrap.ps1"
& $EKSBootstrapScriptFile -EKSClusterName 'test-cluster' -APIServerEndpoint 'https://test-cluster' -Base64ClusterCA 'ca-bundle' -KubeletExtraArgs '--node-labels="karpenter.sh/capacity-type=on-demand,karpenter.sh/nodepool=test" --max-pods=110' -DNSClusterIP '10.100.0.10'
</powershell>

Note

Karpenter will automatically query for the appropriate EKS optimized AMI via AWS Systems Manager (SSM). In the case of the Custom AMIFamily, no default AMIs are defined. As a result, amiSelectorTerms must be specified to inform Karpenter on which custom AMIs are to be used.

Custom

The Custom AMIFamily ships without any default userData to allow you to configure custom bootstrapping for control planes or images that don’t support the default methods from the other families.

spec.subnetSelectorTerms

Subnet Selector Terms allow you to specify selection logic for a set of subnet options that Karpenter can choose from when launching an instance from the EC2NodeClass. Karpenter discovers subnets through the EC2NodeClass using ids or tags. When launching nodes, a subnet is automatically chosen that matches the desired zone. If multiple subnets exist for a zone, the one with the most available IP addresses will be used.

This selection logic is modeled as terms, where each term contains multiple conditions that must all be satisfied for the selector to match. Effectively, all requirements within a single term are ANDed together. It’s possible that you may want to select on two different subnets that have unrelated requirements. In this case, you can specify multiple terms which will be ORed together to form your selection logic. The example below shows how this selection logic is fulfilled.

subnetSelectorTerms:
 # Select on any subnet that has the "karpenter.sh/discovery: ${CLUSTER_NAME}"
 # AND the "environment: test" tag OR any subnet with ID "subnet-09fa4a0a8f233a921"
 - tags:
 karpenter.sh/discovery: "${CLUSTER_NAME}"
 environment: test
 - id: subnet-09fa4a0a8f233a921

Tip

Subnets may be specified by any tag, including Name. Selecting tag values using wildcards (*) is supported.

Examples

Select all with a specified tag key:

spec:
 subnetSelectorTerms:
 - tags:
 karpenter.sh/discovery/MyClusterName: '*'

Select by name and tag (all criteria must match):

spec:
 subnetSelectorTerms:
 - tags:
 Name: my-subnet
 MyTag: '' # matches all resources with the tag

Select using multiple tag terms:

spec:
 subnetSelectorTerms:
 - tags:
 Name: "my-subnet-1"
 - tags:
 Name: "my-subnet-2"

Select using wildcards:

spec:
 subnetSelectorTerms:
 - tags:
 Name: "*Public*"

Select using ids:

spec:
 subnetSelectorTerms:
 - id: "subnet-09fa4a0a8f233a921"
 - id: "subnet-0471ca205b8a129ae"

spec.securityGroupSelectorTerms

Security Group Selector Terms allow you to specify selection logic for all security groups that will be attached to an instance launched from the EC2NodeClass. The security group of an instance is comparable to a set of firewall rules. EKS creates at least two security groups by default.

This selection logic is modeled as terms, where each term contains multiple conditions that must all be satisfied for the selector to match. Effectively, all requirements within a single term are ANDed together. It’s possible that you may want to select on two different security groups that have unrelated requirements. In this case, you can specify multiple terms which will be ORed together to form your selection logic. The example below shows how this selection logic is fulfilled.

securityGroupSelectorTerms:
 # Select on any security group that has both the "karpenter.sh/discovery: ${CLUSTER_NAME}" tag
 # AND the "environment: test" tag OR any security group with the "my-security-group" name
 # OR any security group with ID "sg-063d7acfb4b06c82c"
 - tags:
 karpenter.sh/discovery: "${CLUSTER_NAME}"
 environment: test
 - name: my-security-group
 - id: sg-063d7acfb4b06c82c

Tip

Security groups may be specified by any tag, including “Name”. Selecting tags using wildcards (*) is supported.

Note

When launching nodes, Karpenter uses all the security groups that match the selector. If you choose to use the kubernetes.io/cluster/$CLUSTER_NAME tag for discovery, note that this may result in failures using the AWS Load Balancer controller. The Load Balancer controller only supports a single security group having that tag key. See this issue for more details.

To verify if this restriction affects you, run the following commands.

CLUSTER_VPC_ID="$(aws eks describe-cluster --name $CLUSTER_NAME --query cluster.resourcesVpcConfig.vpcId --output text)"

aws ec2 describe-security-groups --filters Name=vpc-id,Values=$CLUSTER_VPC_ID Name=tag-key,Values=kubernetes.io/cluster/$CLUSTER_NAME --query 'SecurityGroups[].[GroupName]' --output text

If multiple securityGroups are printed, you will need more specific securityGroupSelectorTerms. We generally recommend that you use the karpenter.sh/discovery: $CLUSTER_NAME tag selector instead.

Examples

Select all assigned to a cluster:

spec:
 securityGroupSelectorTerms:
 - tags:
 kubernetes.io/cluster/$CLUSTER_NAME: "owned"

Select all with a specified tag key:

spec:
 securityGroupSelectorTerms:
 - tags:
 MyTag: '*'

Select by name and tag (all criteria must match):

spec:
 securityGroupSelectorTerms:
 - name: my-security-group
 tags:
 MyTag: '*' # matches all resources with the tag

Select using multiple tag terms:

spec:
 securityGroupSelectorTerms:
 - tags:
 Name: "my-security-group-1"
 - tags:
 Name: "my-security-group-2"

Select by name using a wildcard:

spec:
 securityGroupSelectorTerms:
 - name: "*Public*"

Select using ids:

spec:
 securityGroupSelectorTerms:
 - id: "sg-063d7acfb4b06c82c"
 - id: "sg-06e0cf9c198874591"

spec.amiSelectorTerms

AMI Selector Terms are used to configure custom AMIs for Karpenter to use, where the AMIs are discovered through ids, owners, name, and tags. When you specify amiSelectorTerms, you fully override the default AMIs that are selected on by your EC2NodeClass amiFamily.

Note

amiFamily determines the bootstrapping mode, while amiSelectorTerms specifies specific AMIs to be used. Therefore, you need to ensure consistency between amiFamily and amiSelectorTerms to avoid conflicts during bootstrapping.

This selection logic is modeled as terms, where each term contains multiple conditions that must all be satisfied for the selector to match. Effectively, all requirements within a single term are ANDed together. It’s possible that you may want to select on two different AMIs that have unrelated requirements. In this case, you can specify multiple terms which will be ORed together to form your selection logic. The example below shows how this selection logic is fulfilled.

amiSelectorTerms:
 # Select on any AMI that has both the "karpenter.sh/discovery: ${CLUSTER_NAME}" tag
 # AND the "environment: test" tag OR any AMI with the "my-ami" name
 # OR any AMI with ID "ami-123"
 - tags:
 karpenter.sh/discovery: "${CLUSTER_NAME}"
 environment: test
 - name: my-ami
 - id: ami-123

This field is optional, and Karpenter will use the latest EKS-optimized AMIs for the AMIFamily if no amiSelectorTerms are specified. To select an AMI by name, use the name field in the selector term. To select an AMI by id, use the id field in the selector term. To ensure that AMIs are owned by the expected owner, use the owner field - you can use a combination of account aliases (e.g. self amazon, your-aws-account-name) and account IDs.

If owner is not set for name, it defaults to self,amazon, preventing Karpenter from inadvertently selecting an AMI that is owned by a different account. Tags don’t require an owner as tags can only be discovered by the user who created them.

Tip

AMIs may be specified by any AWS tag, including Name. Selecting by tag or by name using wildcards (*) is supported.

Note

If amiSelectorTerms match more than one AMI, Karpenter will automatically determine which AMI best fits the workloads on the launched worker node under the following constraints:

When launching nodes, Karpenter automatically determines which architecture a custom AMI is compatible with and will use images that match an instanceType’s requirements.
- Note that Karpenter cannot detect any requirement other than architecture. If you need to specify different AMIs for different kind of nodes (e.g. accelerated GPU AMIs), you should use a separate EC2NodeClass.
If multiple AMIs are found that can be used, Karpenter will choose the latest one.
If no AMIs are found that can be used, then no nodes will be provisioned.

Examples

Select all with a specified tag:

 amiSelectorTerms:
 - tags:
 karpenter.sh/discovery/MyClusterName: '*'

Select by name:

 amiSelectorTerms:
 - name: my-ami

Select by Name tag:

 amiSelectorTerms:
 - tags:
 Name: my-ami

Select by name and owner:

 amiSelectorTerms:
 - name: my-ami
 owner: self
 - name: my-ami
 owner: "0123456789"

Select by name using a wildcard:

spec:
 amiSelectorTerms:
 - name: "*EKS*"

Select by all under an owner:

spec:
 amiSelectorTerms:
 - name: "*"
 owner: self

Specify using ids:

 amiSelectorTerms:
 - id: "ami-123"
 - id: "ami-456"

spec.role

Role is an optional field and tells Karpenter which IAM identity nodes should assume. You must specify one of role or instanceProfile when creating a Karpenter EC2NodeClass. If using the Karpenter Getting Started Guide to deploy Karpenter, you can use the KarpenterNodeRole-$CLUSTER_NAME role provisioned by that process.

spec:
 role: "KarpenterNodeRole-$CLUSTER_NAME"

spec.instanceProfile

InstanceProfile is an optional field and tells Karpenter which IAM identity nodes should assume. You must specify one of role or instanceProfile when creating a Karpenter EC2NodeClass. If you use the instanceProfile field instead of role, Karpenter will not manage the InstanceProfile on your behalf; instead, it expects that you have pre-provisioned an IAM instance profile and assigned it a role.

You can provision and assign a role to an IAM instance profile using CloudFormation or by using the aws iam create-instance-profile and aws iam add-role-to-instance-profile commands in the CLI.

Note

For private clusters that do not have access to the public internet, using spec.instanceProfile is required. spec.role cannot be used since Karpenter needs to access IAM endpoints to manage a generated instance profile. IAM doesn’t support private endpoints to enable accessing the service without going to the public internet.

spec.tags

Karpenter adds tags to all resources it creates, including EC2 Instances, EBS volumes, and Launch Templates. The default set of tags are listed below.

Name: <node-name>
karpenter.sh/nodeclaim: <nodeclaim-name>
karpenter.sh/nodepool: <nodepool-name>
karpenter.k8s.aws/ec2nodeclass: <ec2nodeclass-name>
kubernetes.io/cluster/<cluster-name>: owned

Additional tags can be added in the tags section, which will be merged with the default tags specified above.

spec:
 tags:
 InternalAccountingTag: 1234
 dev.corp.net/app: Calculator
 dev.corp.net/team: MyTeam

Note

Karpenter allows overrides of the default “Name” tag but does not allow overrides to restricted domains (such as “karpenter.sh”, “karpenter.k8s.aws”, and “kubernetes.io/cluster”). This ensures that Karpenter is able to correctly auto-discover nodes that it owns.

spec.metadataOptions

Control the exposure of Instance Metadata Service on EC2 Instances launched by this EC2NodeClass using a generated launch template.

Refer to recommended, security best practices for limiting exposure of Instance Metadata and User Data to pods.

If metadataOptions are omitted from this EC2NodeClass, the following default settings are applied:

spec:
 metadataOptions:
 httpEndpoint: enabled
 httpProtocolIPv6: disabled
 httpPutResponseHopLimit: 2
 httpTokens: required

spec.blockDeviceMappings

The blockDeviceMappings field in an EC2NodeClass can be used to control the Elastic Block Storage (EBS) volumes that Karpenter attaches to provisioned nodes. Karpenter uses default block device mappings for the AMIFamily specified. For example, the Bottlerocket AMI Family defaults with two block device mappings, one for Bottlerocket’s control volume and the other for container resources such as images and logs.

spec:
 blockDeviceMappings:
 - deviceName: /dev/xvda
 ebs:
 volumeSize: 100Gi
 volumeType: gp3
 iops: 10000
 encrypted: true
 kmsKeyID: "1234abcd-12ab-34cd-56ef-1234567890ab"
 deleteOnTermination: true
 throughput: 125
 snapshotID: snap-0123456789

The following blockDeviceMapping defaults are used for each AMIFamily if no blockDeviceMapping overrides are specified in the EC2NodeClass

AL2

spec:
 blockDeviceMappings:
 - deviceName: /dev/xvda
 ebs:
 volumeSize: 20Gi
 volumeType: gp3
 encrypted: true

Bottlerocket

spec:
 blockDeviceMappings:
 # Root device
 - deviceName: /dev/xvda
 ebs:
 volumeSize: 4Gi
 volumeType: gp3
 encrypted: true
 # Data device: Container resources such as images and logs
 - deviceName: /dev/xvdb
 ebs:
 volumeSize: 20Gi
 volumeType: gp3
 encrypted: true

Ubuntu

spec:
 blockDeviceMappings:
 - deviceName: /dev/sda1
 ebs:
 volumeSize: 20Gi
 volumeType: gp3
 encrypted: true

Windows2019/Windows2022

spec:
 blockDeviceMappings:
 - deviceName: /dev/sda1
 ebs:
 volumeSize: 50Gi
 volumeType: gp3
 encrypted: true

Custom

The Custom AMIFamily ships without any default blockDeviceMappings.

spec.userData

You can control the UserData that is applied to your worker nodes via this field. This allows you to run custom scripts or pass-through custom configuration to Karpenter instances on start-up.

apiVersion: karpenter.k8s.aws/v1beta1
kind: EC2NodeClass
metadata:
 name: bottlerocket-example
spec:
 ...
 amiFamily: Bottlerocket
 userData: |
 [settings.kubernetes]
 "kube-api-qps" = 30
 "shutdown-grace-period" = "30s"
 "shutdown-grace-period-for-critical-pods" = "30s"
 [settings.kubernetes.eviction-hard]
 "memory.available" = "20%"

This example adds SSH keys to allow remote login to the node (replace my-authorized_keys with your key file):

Note

Instead of using SSH as set up in this example, you can use Session Manager (SSM) or EC2 Instance Connect to gain shell access to Karpenter nodes. See Node NotReady troubleshooting for an example of starting an SSM session from the command line or EC2 Instance Connect documentation to connect to nodes using SSH.

apiVersion: karpenter.k8s.aws/v1beta1
kind: EC2NodeClass
metadata:
 name: al2-example
spec:
 ...
 amiFamily: AL2
 userData: |
 #!/bin/bash
 mkdir -p ~ec2-user/.ssh/
 touch ~ec2-user/.ssh/authorized_keys
 cat >> ~ec2-user/.ssh/authorized_keys <<EOF
 {{ insertFile "../my-authorized_keys" | indent 4 }}
 EOF
 chmod -R go-w ~ec2-user/.ssh/authorized_keys
 chown -R ec2-user ~ec2-user/.ssh

For more examples on configuring fields for different AMI families, see the examples here.

Karpenter will merge the userData you specify with the default userData for that AMIFamily. See the AMIFamily section for more details on these defaults. View the sections below to understand the different merge strategies for each AMIFamily.

AL2/Ubuntu

Your UserData can be in the MIME multi part archive format.
Karpenter will transform your custom user-data as a MIME part, if necessary, and then merge a final MIME part to the end of your UserData parts which will bootstrap the worker node. Karpenter will have full control over all the parameters being passed to the bootstrap script.
- Karpenter will continue to set MaxPods, ClusterDNS and all other parameters defined in spec.kubeletConfiguration as before.

Consider the following example to understand how your custom UserData will be merged -

Passed-in UserData (bash)

#!/bin/bash
echo "Running custom user data script"

Passed-in UserData (MIME)

MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="BOUNDARY"

--BOUNDARY
Content-Type: text/x-shellscript; charset="us-ascii"

#!/bin/bash
echo "Running custom user data script"

--BOUNDARY--

Merged UserData

MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="//"

--//
Content-Type: text/x-shellscript; charset="us-ascii"

#!/bin/bash
echo "Running custom user data script"

--//
Content-Type: text/x-shellscript; charset="us-ascii"

#!/bin/bash -xe
exec > >(tee /var/log/user-data.log|logger -t user-data -s 2>/dev/console) 2>&1
/etc/eks/bootstrap.sh 'test-cluster' --apiserver-endpoint 'https://test-cluster' --b64-cluster-ca 'ca-bundle' \
--use-max-pods false \
--container-runtime containerd \
--kubelet-extra-args '--node-labels=karpenter.sh/capacity-type=on-demand,karpenter.sh/nodepool=test --max-pods=110'
--//--

Note

You can also set kubelet-config properties by modifying the kubelet-config.json file before the EKS bootstrap script starts the kubelet:

apiVersion: karpenter.k8s.aws/v1beta1
kind: EC2NodeClass
metadata:
 name: kubelet-config-example
spec:
 ...
 amiFamily: AL2
 userData: |
 #!/bin/bash
 echo "$(jq '.kubeAPIQPS=50' /etc/kubernetes/kubelet/kubelet-config.json)" > /etc/kubernetes/kubelet/kubelet-config.json

Bottlerocket

Your UserData must be valid TOML.
Karpenter will automatically merge settings to ensure successful bootstrap including cluster-name, api-server and cluster-certificate. Any labels and taints that need to be set based on pod requirements will also be specified in the final merged UserData.
- All Kubelet settings that Karpenter applies will override the corresponding settings in the provided UserData. For example, if you’ve specified settings.kubernetes.cluster-name, it will be overridden.
- If MaxPods is specified via the binary arg to Karpenter, the value will override anything specified in the UserData.
- If ClusterDNS is specified via spec.kubeletConfiguration, then that value will override anything specified in the UserData.
Unknown TOML fields will be ignored when the final merged UserData is generated by Karpenter.

Consider the following example to understand how your custom UserData settings will be merged in.

Passed-in UserData

[settings.kubernetes.eviction-hard]
"memory.available" = "12%"
[settings.kubernetes]
"unknown-setting" = "unknown"
[settings.kubernetes.node-labels]
'field.controlled.by/karpenter' = 'will-be-overridden'

Merged UserData

[settings]
[settings.kubernetes]
api-server = 'https://cluster'
cluster-certificate = 'ca-bundle'
cluster-name = 'cluster'

[settings.kubernetes.node-labels]
'karpenter.sh/capacity-type' = 'on-demand'
'karpenter.sh/nodepool' = 'default'

[settings.kubernetes.node-taints]

[settings.kubernetes.eviction-hard]
'memory.available' = '12%%'

Windows2019/Windows2022

Your UserData must be specified as PowerShell commands.
The UserData specified will be prepended to a Karpenter managed section that will bootstrap the kubelet.
Karpenter will continue to set ClusterDNS and all other parameters defined in spec.kubeletConfiguration as before.

Consider the following example to understand how your custom UserData settings will be merged in.

Passed-in UserData

Write-Host "Running custom user data script"

Merged UserData

<powershell>
Write-Host "Running custom user data script"
[string]$EKSBootstrapScriptFile = "$env:ProgramFiles\Amazon\EKS\Start-EKSBootstrap.ps1"
& $EKSBootstrapScriptFile -EKSClusterName 'test-cluster' -APIServerEndpoint 'https://test-cluster' -Base64ClusterCA 'ca-bundle' -KubeletExtraArgs '--node-labels="karpenter.sh/capacity-type=spot,karpenter.sh/nodepool=windows2022" --max-pods=110' -DNSClusterIP '10.0.100.10'
</powershell>

Windows Support Notice

Currently, Karpenter does not specify -ServiceCIDR to EKS Windows AMI Bootstrap script. Windows worker nodes will use 172.20.0.0/16 or 10.100.0.0/16 for Kubernetes service IP address ranges based on the IP address of the primary interface. The effective ServiceCIDR can be verified at $env:ProgramData\Amazon\EKS\cni\config\vpc-bridge.conf on the worker node.

Support for the Windows ServiceCIDR argument can be tracked in a Karpenter Github Issue. Currently, if the effective ServiceCIDR is incorrect for your windows worker nodes, you can add the following userData as a workaround.

spec:
 userData: |
 $global:EKSCluster = Get-EKSCluster -Name my-cluster

Custom

No merging is performed, your UserData must perform all setup required of the node to allow it to join the cluster.

spec.detailedMonitoring

Enabling detailed monitoring controls the EC2 detailed monitoring feature. If you enable this option, the Amazon EC2 console displays monitoring graphs with a 1-minute period for the instances that Karpenter launches.

spec:
 detailedMonitoring: true

status.subnets

status.subnets contains the resolved id and zone of the subnets that were selected by the spec.subnetSelectorTerms for the node class. The subnets will be sorted by the available IP address count in decreasing order.

Examples

spec:
 subnetSelectorTerms:
 - tags:
 karpenter.sh/discovery: "${CLUSTER_NAME}"
status:
 subnets:
 - id: subnet-0a462d98193ff9fac
 zone: us-east-2b
 - id: subnet-0322dfafd76a609b6
 zone: us-east-2c
 - id: subnet-0727ef01daf4ac9fe
 zone: us-east-2b
 - id: subnet-00c99aeafe2a70304
 zone: us-east-2a
 - id: subnet-023b232fd5eb0028e
 zone: us-east-2c
 - id: subnet-03941e7ad6afeaa72
 zone: us-east-2a

status.securityGroups

status.securityGroups contains the resolved id and name of the security groups that were selected by the spec.securityGroupSelectorTerms for the node class. The subnets will be sorted by the available IP address count in decreasing order.

Examples

spec:
 securityGroupSelectorTerms:
 - tags:
 karpenter.sh/discovery: "${CLUSTER_NAME}"
status:
 securityGroups:
 - id: sg-041513b454818610b
 name: ClusterSharedNodeSecurityGroup
 - id: sg-0286715698b894bca
 name: ControlPlaneSecurityGroup-1AQ073TSAAPW

status.amis

status.amis contains the resolved id, name, and requirements of either the default AMIs for the spec.amiFamily or the AMIs selected by the spec.amiSelectorTerms if this field is specified.

Examples

Default AMIs resolved from the AL2 AMIFamily:

spec:
 amiFamily: AL2
status:
 amis:
 - id: ami-03c3a3dcda64f5b75
 name: amazon-linux-2-gpu
 requirements:
 - key: kubernetes.io/arch
 operator: In
 values:
 - amd64
 - key: karpenter.k8s.aws/instance-gpu-count
 operator: Exists
 - id: ami-03c3a3dcda64f5b75
 name: amazon-linux-2-gpu
 requirements:
 - key: kubernetes.io/arch
 operator: In
 values:
 - amd64
 - key: karpenter.k8s.aws/instance-accelerator-count
 operator: Exists
 - id: ami-06afb2d101cc4b8bd
 name: amazon-linux-2-arm64
 requirements:
 - key: kubernetes.io/arch
 operator: In
 values:
 - arm64
 - key: karpenter.k8s.aws/instance-gpu-count
 operator: DoesNotExist
 - key: karpenter.k8s.aws/instance-accelerator-count
 operator: DoesNotExist
 - id: ami-0e28b76d768af234e
 name: amazon-linux-2
 requirements:
 - key: kubernetes.io/arch
 operator: In
 values:
 - amd64
 - key: karpenter.k8s.aws/instance-gpu-count
 operator: DoesNotExist
 - key: karpenter.k8s.aws/instance-accelerator-count
 operator: DoesNotExist

AMIs resolved from spec.amiSelectorTerms:

spec:
 amiFamily: AL2
 amiSelectorTerms:
 - tags:
 karpenter.sh/discovery: "${CLUSTER_NAME}"
status:
 amis:
 - id: ami-01234567890123456
 name: custom-ami-amd64
 requirements:
 - key: kubernetes.io/arch
 operator: In
 values:
 - amd64
 - id: ami-01234567890123456
 name: custom-ami-arm64
 requirements:
 - key: kubernetes.io/arch
 operator: In
 values:
 - arm64

status.instanceProfile

status.instanceProfile contains the resolved instance profile generated by Karpenter from the spec.role

spec:
 role: "KarpenterNodeRole-${CLUSTER_NAME}"
status:
 instanceProfile: "${CLUSTER_NAME}-0123456778901234567789"

V0.32: Scheduling

Mon, 01 Jan 0001 00:00:00 +0000

If your pods have no requirements for how or where to run, you can let Karpenter choose nodes from the full range of available cloud provider resources. However, by taking advantage of Karpenter’s model of layered constraints, you can be sure that the precise type and amount of resources needed are available to your pods. Reasons for constraining where your pods run could include:

Needing to run in zones where dependent applications or storage are available
Requiring certain kinds of processors or other hardware
Wanting to use techniques like topology spread to help ensure high availability

Your Cloud Provider defines the first layer of constraints, including all instance types, architectures, zones, and purchase types available to its cloud. The cluster administrator adds the next layer of constraints by creating one or more NodePools. The final layer comes from you adding specifications to your Kubernetes pod deployments. Pod scheduling constraints must fall within a NodePool’s constraints or the pods will not deploy. For example, if the NodePool sets limits that allow only a particular zone to be used, and a pod asks for a different zone, it will not be scheduled.

Constraints you can request include:

Resource requests: Request that certain amount of memory or CPU be available.
Node selection: Choose to run on a node that is has a particular label (nodeSelector).
Node affinity: Draws a pod to run on nodes with particular attributes (affinity).
Topology spread: Use topology spread to help ensure availability of the application.
Pod affinity/anti-affinity: Draws pods towards or away from topology domains based on the scheduling of other pods.

Karpenter supports standard Kubernetes scheduling constraints. This allows you to define a single set of rules that apply to both existing and provisioned capacity.

Note

Karpenter supports specific Well-Known Labels, Annotations and Taints that are useful for scheduling.

Resource requests

Within a Pod spec, you can both make requests and set limits on resources a pod needs, such as CPU and memory. For example:

apiVersion: v1
kind: Pod
metadata:
 name: myapp
spec:
 containers:
 - name: app
 image: myimage
 resources:
 requests:
 memory: "128Mi"
 cpu: "500m"
 limits:
 memory: "256Mi"
 cpu: "1000m"

In this example, the container is requesting 128MiB of memory and .5 CPU. Its limits are set to 256MiB of memory and 1 CPU. Instance type selection math only uses requests, but limits may be configured to enable resource oversubscription.

See Managing Resources for Containers for details on resource types supported by Kubernetes, Specify a memory request and a memory limit for examples of memory requests, and NodePools for a list of supported resources.

Accelerators/GPU Resources

Accelerator (e.g., GPU) values include

nvidia.com/gpu
amd.com/gpu
aws.amazon.com/neuron
habana.ai/gaudi

Karpenter supports accelerators, such as GPUs.

Additionally, include a resource requirement in the workload manifest. This will cause the GPU dependent pod to be scheduled onto the appropriate node.

Here is an example of an accelerator resource in a workload manifest (e.g., pod):

spec:
 template:
 spec:
 containers:
 - resources:
 limits:
 nvidia.com/gpu: "1"

Note

If you are provisioning GPU nodes, you need to deploy an appropriate GPU device plugin daemonset for those nodes. Without the daemonset running, Karpenter will not see those nodes as initialized. Refer to general Kubernetes GPU docs and the following specific GPU docs:

nvidia.com/gpu: NVIDIA device plugin for Kubernetes
amd.com/gpu: AMD GPU device plugin for Kubernetes
aws.amazon.com/neuron: Kubernetes environment setup for Neuron
habana.ai/gaudi: Habana device plugin for Kubernetes

Pod ENI Resources (Security Groups for Pods)

Pod ENI is a feature of the AWS VPC CNI Plugin which allows an Elastic Network Interface (ENI) to be allocated directly to a Pod. When enabled, the vpc.amazonaws.com/pod-eni extended resource is added to supported nodes. The Pod ENI feature can be used independently, but is most often used in conjunction with Security Groups for Pods. Follow the below instructions to enable support for Pod ENI and/or Security Groups for Pods in Karpenter.

Note

You must enable Pod ENI support in the AWS VPC CNI Plugin before enabling Pod ENI support in Karpenter. Please refer to the Security Groups for Pods documentation for instructions.

Note

If you’ve enabled Security Groups per Pod, one of the instance’s ENIs is reserved. To avoid discrepancies between the maxPods value and the node’s supported pod density, you need to set RESERVED_ENIS=1.

Here is an example of a pod-eni resource defined in a deployment manifest:

spec:
template:
spec:
containers:
- resources:
limits:
vpc.amazonaws.com/pod-eni: "1"

Windows Support Notice

Security groups for pods are currently unsupported for Windows nodes

Selecting nodes

With nodeSelector you can ask for a node that matches selected key-value pairs. This can include well-known labels or custom labels you create yourself.

You can use affinity to define more complicated constraints, see Node Affinity for the complete specification.

Labels

Well-known labels may be specified as NodePool requirements or pod scheduling constraints. You can also define your own custom labels by specifying requirements or labels on your NodePool and select them using nodeAffinity or nodeSelectors on your Pods.

Warning

Take care to ensure the label domains are correct. A well known label like karpenter.k8s.aws/instance-family will enforce node properties, but may be confused with node.kubernetes.io/instance-family, which is unknown to Karpenter, and treated as a custom label which will not enforce node properties.

Well-Known Labels

Label	Example	Description
topology.kubernetes.io/zone	us-east-2a	Zones are defined by your cloud provider (aws)
node.kubernetes.io/instance-type	g4dn.8xlarge	Instance types are defined by your cloud provider (aws)
node.kubernetes.io/windows-build	10.0.17763	Windows OS build in the format “MajorVersion.MinorVersion.BuildNumber”. Can be `10.0.17763` for WS2019, or `10.0.20348` for WS2022. (k8s)
kubernetes.io/os	linux	Operating systems are defined by GOOS values on the instance
kubernetes.io/arch	amd64	Architectures are defined by GOARCH values on the instance
karpenter.sh/capacity-type	spot	Capacity types include `spot`, `on-demand`
karpenter.k8s.aws/instance-hypervisor	nitro	[AWS Specific] Instance types that use a specific hypervisor
karpenter.k8s.aws/instance-encryption-in-transit-supported	true	[AWS Specific] Instance types that support (or not) in-transit encryption
karpenter.k8s.aws/instance-category	g	[AWS Specific] Instance types of the same category, usually the string before the generation number
karpenter.k8s.aws/instance-generation	4	[AWS Specific] Instance type generation number within an instance category
karpenter.k8s.aws/instance-family	g4dn	[AWS Specific] Instance types of similar properties but different resource quantities
karpenter.k8s.aws/instance-size	8xlarge	[AWS Specific] Instance types of similar resource quantities but different properties
karpenter.k8s.aws/instance-cpu	32	[AWS Specific] Number of CPUs on the instance
karpenter.k8s.aws/instance-memory	131072	[AWS Specific] Number of mebibytes of memory on the instance
karpenter.k8s.aws/instance-network-bandwidth	131072	[AWS Specific] Number of baseline megabits available on the instance
karpenter.k8s.aws/instance-gpu-name	t4	[AWS Specific] Name of the GPU on the instance, if available
karpenter.k8s.aws/instance-gpu-manufacturer	nvidia	[AWS Specific] Name of the GPU manufacturer
karpenter.k8s.aws/instance-gpu-count	1	[AWS Specific] Number of GPUs on the instance
karpenter.k8s.aws/instance-gpu-memory	16384	[AWS Specific] Number of mebibytes of memory on the GPU
karpenter.k8s.aws/instance-local-nvme	900	[AWS Specific] Number of gibibytes of local nvme storage on the instance

Note

Karpenter translates the following deprecated labels to their stable equivalents: failure-domain.beta.kubernetes.io/zone, failure-domain.beta.kubernetes.io/region, beta.kubernetes.io/arch, beta.kubernetes.io/os, and beta.kubernetes.io/instance-type.

User-Defined Labels

Karpenter is aware of several well-known labels, deriving them from instance type details. If you specify a nodeSelector or a required nodeAffinity using a label that is not well-known to Karpenter, it will not launch nodes with these labels and pods will remain pending. For Karpenter to become aware that it can schedule for these labels, you must specify the label in the NodePool requirements with the Exists operator:

requirements:
 - key: user.defined.label/type
 operator: Exists

Note

Node selectors

Here is an example of a nodeSelector for selecting nodes:

nodeSelector:
 topology.kubernetes.io/zone: us-west-2a
 karpenter.sh/capacity-type: spot

This example features a well-known label (topology.kubernetes.io/zone) and a label that is well known to Karpenter (karpenter.sh/capacity-type).

If you want to create a custom label, you should do that at the NodePool level. Then the pod can declare that custom label.

See nodeSelector in the Kubernetes documentation for details.

Node affinity

Examples below illustrate how to use Node affinity to include (In) and exclude (NotIn) objects. See Node affinity for details. When setting rules, the following Node affinity types define how hard or soft each rule is:

requiredDuringSchedulingIgnoredDuringExecution: This is a hard rule that must be met.
preferredDuringSchedulingIgnoredDuringExecution: This is a preference, but the pod can run on a node where it is not guaranteed.

The IgnoredDuringExecution part of each tells the pod to keep running, even if conditions change on the node so the rules no longer matched. You can think of these concepts as required and preferred, since Kubernetes never implemented other variants of these rules.

All examples below assume that the NodePool doesn’t have constraints to prevent those zones from being used. The first constraint says you could use us-west-2a or us-west-2b, the second constraint makes it so only us-west-2b can be used.

 affinity:
 nodeAffinity:
 requiredDuringSchedulingIgnoredDuringExecution:
 nodeSelectorTerms:
 - matchExpressions:
 - key: "topology.kubernetes.io/zone"
 operator: "In"
 values: ["us-west-2a", "us-west-2b"]
 - key: "topology.kubernetes.io/zone"
 operator: "In"
 values: ["us-west-2b"]

Changing the second operator to NotIn would allow the pod to run in us-west-2a only:

 - key: "topology.kubernetes.io/zone"
 operator: "In"
 values: ["us-west-2a", "us-west-2b"]
 - key: "topology.kubernetes.io/zone"
 operator: "NotIn"
 values: ["us-west-2b"]

Continuing to add to the example, nodeAffinity lets you define terms so if one term doesn’t work it goes to the next one. Here, if us-west-2a is not available, the second term will cause the pod to run on a spot instance in us-west-2d.

 affinity:
 nodeAffinity:
 requiredDuringSchedulingIgnoredDuringExecution:
 nodeSelectorTerms:
 - matchExpressions: # OR
 - key: "topology.kubernetes.io/zone" # AND
 operator: "In"
 values: ["us-west-2a", "us-west-2b"]
 - key: "topology.kubernetes.io/zone" # AND
 operator: "NotIn"
 values: ["us-west-2b"]
 - matchExpressions: # OR
 - key: "karpenter.sh/capacity-type" # AND
 operator: "In"
 values: ["spot"]
 - key: "topology.kubernetes.io/zone" # AND
 operator: "In"
 values: ["us-west-2d"]

In general, Karpenter will go through each of the nodeSelectorTerms in order and take the first one that works. However, if Karpenter fails to provision on the first nodeSelectorTerms, it will try again using the second one. If they all fail, Karpenter will fail to provision the pod. Karpenter will backoff and retry over time. So if capacity becomes available, it will schedule the pod without user intervention.

Taints and tolerations

Taints are the opposite of affinity. Setting a taint on a node tells the scheduler to not run a pod on it unless the pod has explicitly said it can tolerate that taint. This example shows a NodePool that was set up with a taint for only running pods that require a GPU, such as the following:

apiVersion: karpenter.sh/v1beta1
kind: NodePool
metadata:
 name: gpu
spec:
 template:
 spec:
 requirements:
 - key: karpenter.k8s.aws/instance-family
 operator: In
 values:
 - p3
 taints:
 - key: nvidia.com/gpu
 value: "true"
 effect: "NoSchedule"

For a pod to request to run on a node that has this NodePool, it could set a toleration as follows:

apiVersion: v1
kind: Pod
metadata:
 name: mygpupod
spec:
 containers:
 - name: gpuapp
 resources:
 requests:
 nvidia.com/gpu: 1
 limits:
 nvidia.com/gpu: 1
 image: mygpucontainer
 tolerations:
 - key: "nvidia.com/gpu"
 operator: "Exists"
 effect: "NoSchedule"

See Taints and Tolerations in the Kubernetes documentation for details.

Topology Spread

By using the Kubernetes topologySpreadConstraints you can ask the NodePool to have pods push away from each other to limit the blast radius of an outage. Think of it as the Kubernetes evolution for pod affinity: it lets you relate pods with respect to nodes while still allowing spread. For example:

spec:
 topologySpreadConstraints:
 - maxSkew: 1
 topologyKey: "topology.kubernetes.io/zone"
 whenUnsatisfiable: ScheduleAnyway
 labelSelector:
 matchLabels:
 dev: jjones
 - maxSkew: 1
 topologyKey: "kubernetes.io/hostname"
 whenUnsatisfiable: ScheduleAnyway
 labelSelector:
 matchLabels:
 dev: jjones
 - maxSkew: 1
 topologyKey: "karpenter.sh/capacity-type"
 whenUnsatisfiable: ScheduleAnyway
 labelSelector:
 matchLabels:
 dev: jjones

Adding this to your podspec would result in:

Pods being spread across zones, hosts, and capacity-type (topologyKey).
The dev labelSelector will include all pods with the label of dev=jjones in topology calculations. It is recommended to use a selector to match all pods in a deployment.
No more than one pod difference in the number of pods on each host (maxSkew). For example, if there were three nodes and five pods the pods could be spread 1, 2, 2 or 2, 1, 2 and so on. If instead the spread were 5, pods could be 5, 0, 0 or 3, 2, 0, or 2, 1, 2 and so on.

The three supported topologyKey values that Karpenter supports are:

topology.kubernetes.io/zone
kubernetes.io/hostname
karpenter.sh/capacity-type

See Pod Topology Spread Constraints for details.

Note

NodePools do not attempt to balance or rebalance the availability zones for their nodes. Availability zone balancing may be achieved by defining zonal Topology Spread Constraints for Pods that require multi-zone durability, and NodePools will respect these constraints while optimizing for compute costs.

Pod affinity/anti-affinity

By using the podAffinity and podAntiAffinity configuration on a pod spec, you can inform the Karpenter scheduler of your desire for pods to schedule together or apart with respect to different topology domains. For example:

spec:
 affinity:
 podAffinity:
 requiredDuringSchedulingIgnoredDuringExecution:
 - labelSelector:
 matchExpressions:
 - key: system
 operator: In
 values:
 - backend
 topologyKey: topology.kubernetes.io/zone
 podAntiAffinity:
 requiredDuringSchedulingIgnoredDuringExecution:
 - labelSelector:
 matchLabels:
 app: inflate
 topologyKey: kubernetes.io/hostname

The above pod affinity rule would cause the pod to only schedule in zones where a pod with the label system=backend is already running.

The anti-affinity rule would cause it to avoid running on any node with a pod labeled app=inflate. If this anti-affinity term was on a deployment pod spec along with a matching app=inflate label, it would prevent more than one pod from the deployment from running on any single node.

See Inter-pod affinity and anti-affinity in the Kubernetes documentation for details.

Persistent Volume Topology

Karpenter automatically detects storage scheduling requirements and includes them in node launch decisions.

In the following example, the StorageClass defines zonal topologies for us-west-2a and us-west-2b and binding mode WaitForFirstConsumer. When the pod is created, Karpenter follows references from the Pod to PersistentVolumeClaim to StorageClass and identifies that this pod requires storage in us-west-2a and us-west-2b. It randomly selects us-west-2a, provisions a node in that zone, and waits for kube-scheduler to bind the pod to the node. The CSI driver creates a PersistentVolume according to the PersistentVolumeClaim and gives it a node affinity rule for us-west-2a.

Later on, the pod is deleted and a new pod is created that requests the same claim. This time, Karpenter identifies that a PersistentVolume already exists for the PersistentVolumeClaim, and includes its zone us-west-2a in the pod’s scheduling requirements.

apiVersion: v1
kind: Pod
metadata:
 name: app
spec:
 containers: ...
 volumes:
 - name: storage
 persistentVolumeClaim:
 claimName: ebs-claim
---
kind: StorageClass
apiVersion: storage.k8s.io/v1
metadata:
 name: ebs
provisioner: ebs.csi.aws.com
volumeBindingMode: WaitForFirstConsumer
allowedTopologies:
- matchLabelExpressions:
 - key: topology.ebs.csi.aws.com/zone
 values: ["us-west-2a", "us-west-2b"]
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
 name: ebs-claim
spec:
 accessModes:
 - ReadWriteOnce
 storageClassName: ebs
 resources:
 requests:
 storage: 4Gi

Note

☁️ AWS Specific

The EBS CSI driver uses topology.ebs.csi.aws.com/zone instead of the standard topology.kubernetes.io/zone label. Karpenter is aware of label aliasing and translates this label into topology.kubernetes.io/zone in memory. When configuring a StorageClass for the EBS CSI Driver, you must use topology.ebs.csi.aws.com/zone.

Note

The topology key topology.kubernetes.io/region is not supported. Legacy in-tree CSI providers specify this label. Instead, install an out-of-tree CSI provider. Learn more about moving to CSI providers.

Weighted NodePools

Karpenter allows you to order your NodePools using the .spec.weight field so that the Karpenter scheduler will attempt to schedule one NodePool before another.

Savings Plans and Reserved Instances

If you have purchased a Savings Plan or Reserved Instances, you may want to tell Karpenter to prioritize this reserved capacity ahead of other instance types.

To enable this, you will need to tell the Karpenter controllers which instance types to prioritize and what is the maximum amount of capacity that should be provisioned using those instance types. We can set the .spec.limits field on the NodePool to limit the capacity that can be launched by this NodePool. Combined with the .spec.weight value, we can tell Karpenter to pull from instance types in the reserved NodePool before defaulting to generic instance types.

apiVersion: karpenter.sh/v1beta1
kind: NodePool
metadata:
 name: reserved-instance
spec:
 weight: 50
 limits:
 cpu: 100
 template:
 spec:
 requirements:
 - key: "node.kubernetes.io/instance-type"
 operator: In
 values: ["c4.large"]
---
apiVersion: karpenter.sh/v1beta1
kind: NodePool
metadata:
 name: default
spec:
 template:
 spec:
 requirements:
 - key: karpenter.sh/capacity-type
 operator: In
 values: ["spot", "on-demand"]
 - key: kubernetes.io/arch
 operator: In
 values: ["amd64"]

Fallback

Pods that do not specify node selectors or affinities can potentially be assigned to any node with any configuration. There may be cases where you require these pods to schedule to a specific capacity type or architecture but assigning the relevant node selectors or affinities to all these workload pods may be too tedious or infeasible. Instead, we want to define a cluster-wide default configuration for nodes launched using Karpenter.

By assigning a higher .spec.weight value and restricting a NodePool to a specific capacity type or architecture, we can set default configuration for the nodes launched by pods that don’t have node configuration restrictions.

apiVersion: karpenter.sh/v1beta1
kind: NodePool
metadata:
 name: default
spec:
 weight: 50
 template:
 spec:
 requirements:
 - key: karpenter.sh/capacity-type
 operator: In
 values: ["spot", "on-demand"]
 - key: kubernetes.io/arch
 operator: In
 values: ["amd64"]
---
apiVersion: karpenter.sh/v1beta1
kind: NodePool
metadata:
 name: arm64-specific
spec:
 template:
 spec:
 requirements:
 - key: karpenter.sh/capacity-type
 operator: In
 values: ["spot", "on-demand"]
 - key: kubernetes.io/arch
 operator: In
 values: ["arm64"]
 - key: node.kubernetes.io/instance-type
 operator: In
 values: ["a1.large", "a1.xlarge"]

Note

Based on the way that Karpenter performs pod batching and bin packing, it is not guaranteed that Karpenter will always choose the highest priority NodePool given specific requirements. For example, if a pod can’t be scheduled with the highest priority NodePool, it will force creation of a node using a lower priority NodePool, allowing other pods from that batch to also schedule on that node. The behavior may also occur if existing capacity is available, as the kube-scheduler will schedule the pods instead of allowing Karpenter to provision a new node.

Advanced Scheduling Techniques

Scheduling based on Node Resources

You may want pods to be able to request resources of nodes that Kubernetes natively does not provide as a schedulable resource or that are aspects of certain nodes like High Performance Networking or NVME Local Storage. You can use Karpenter’s Well-Known Labels to accomplish this.

These can further be applied at the NodePool or Workload level using Requirements, NodeSelectors or Affinities

Pod example of requiring any NVME disk:

...
 affinity:
 nodeAffinity:
 requiredDuringSchedulingIgnoredDuringExecution:
 nodeSelectorTerms:
 - matchExpressions:
 - key: "karpenter.k8s.aws/instance-local-nvme"
 operator: "Exists"
...

NodePool Example:

...
requirement:
 - key: "karpenter.k8s.aws/instance-local-nvme"
 operator: "Exists"
...

Pod example of requiring at least 100GB of NVME disk:

...
 affinity:
 nodeAffinity:
 requiredDuringSchedulingIgnoredDuringExecution:
 nodeSelectorTerms:
 - matchExpressions:
 - key: "karpenter.k8s.aws/instance-local-nvme"
 operator: Gt
 values: ["99"]
...

NodePool Example:

...
requirement:
 - key: "karpenter.k8s.aws/instance-local-nvme"
 operator: Gt
 values: ["99"]
...

Note

Karpenter cannot yet take into account ephemeral-storage requests while scheduling pods, we’re purely requesting attributes of nodes and getting X amount of resources as a side effect. You may need to tweak schedulable resources like CPU or Memory to achieve desired fit, especially if Consolidation is enabled.

Your NodeClass will also need to support automatically formatting and mounting NVME Instance Storage if available.

Pod example of requiring at least 50 Gbps of network bandwidth:

...
 affinity:
 nodeAffinity:
 requiredDuringSchedulingIgnoredDuringExecution:
 nodeSelectorTerms:
 - matchExpressions:
 - key: "karpenter.k8s.aws/instance-network-bandwidth"
 operator: Gt
 values: ["49999"]
...

NodePool Example:

...
requirement:
 - key: "karpenter.k8s.aws/instance-network-bandwidth"
 operator: Gt
 values: ["49999"]
...

Note

If using Gt/Lt operators, make sure to use values under the actual label values of the desired resource.

`Exists` Operator

The Exists operator can be used on a NodePool to provide workload segregation across nodes.

apiVersion: karpenter.sh/v1beta1
kind: NodePool
spec:
 template:
 spec:
 requirements:
 - key: company.com/team
 operator: Exists
...

With this requirement on the NodePool, workloads can specify the same key (e.g. company.com/team) with custom values (e.g. team-a, team-b, etc.) as a required nodeAffinity or nodeSelector. Karpenter will then apply the key/value pair to nodes it launches dynamically based on the pod’s node requirements.

If each set of pods that can schedule with this NodePool specifies this key in its nodeAffinity or nodeSelector, you can isolate pods onto different nodes based on their values. This provides a way to more dynamically isolate workloads without requiring a unique NodePool for each workload subset.

For example, providing the following nodeSelectors would isolate the pods for each of these deployments on different nodes.

Team A Deployment

apiVersion: v1
kind: Deployment
metadata:
 name: team-a-deployment
spec:
 replicas: 5
 template:
 spec:
 nodeSelector:
 company.com/team: team-a

Team A Node

apiVersion: v1
kind: Node
metadata:
 labels:
 company.com/team: team-a

Team B Deployment

apiVersion: apps/v1
kind: Deployment
metadata:
 name: team-b-deployment
spec:
 replicas: 5
 template:
 spec:
 nodeSelector:
 company.com/team: team-b

Team B Node

apiVersion: v1
kind: Node
metadata:
 labels:
 company.com/team: team-b

Note

If a workload matches the NodePool but doesn’t specify a label, Karpenter will generate a random label for the node.

On-Demand/Spot Ratio Split

Taking advantage of Karpenter’s ability to assign labels to node and using a topology spread across those labels enables a crude method for splitting a workload across on-demand and spot instances in a desired ratio.

To do this, we create one NodePool each for spot and on-demand with disjoint values for a unique new label called capacity-spread. In the example below, we provide four unique values for the spot NodePool and one value for the on-demand NodePool. When we spread across our new label evenly, we’ll end up with a ratio of 4:1 spot to on-demand nodes.

Warning

This is not identical to a topology spread with a specified ratio. We are constructing ‘virtual domains’ to spread evenly across and the ratio of those ‘virtual domains’ to spot and on-demand happen to coincide with the desired spot to on-demand ratio. As an example, if you launch pods using the provided example, Karpenter will launch nodes with capacity-spread labels of 1, 2, 3, 4, and 5. kube-scheduler will then schedule evenly across those nodes to give the desired ratio.

NodePools

apiVersion: karpenter.sh/v1beta1
kind: NodePool
metadata:
 name: spot
spec:
 template:
 spec:
 requirements:
 - key: "karpenter.sh/capacity-type"
 operator: In
 values: ["spot"]
 - key: capacity-spread
 operator: In
 values:
 - "2"
 - "3"
 - "4"
 - "5"
---
apiVersion: karpenter.sh/v1beta1
kind: NodePool
metadata:
 name: on-demand
spec:
 template:
 spec:
 requirements:
 - key: "karpenter.sh/capacity-type"
 operator: In
 values: ["on-demand"]
 - key: capacity-spread
 operator: In
 values:
 - "1"

Workload Topology Spread Constraint

topologySpreadConstraints:
- maxSkew: 1
 topologyKey: capacity-spread
 whenUnsatisfiable: DoNotSchedule
 labelSelector:
 ...

V0.32: Disruption

Mon, 01 Jan 0001 00:00:00 +0000

Control Flow

Karpenter sets a Kubernetes finalizer on each node and node claim it provisions. The finalizer blocks deletion of the node object while the Termination Controller taints and drains the node, before removing the underlying NodeClaim. Disruption is triggered by the Disruption Controller, by the user through manual disruption, or through an external system that sends a delete request to the node object.

Disruption Controller

Karpenter automatically discovers disruptable nodes and spins up replacements when needed. Karpenter disrupts nodes by executing one automated method at a time, in order of Expiration, Drift, and then Consolidation. Each method varies slightly, but they all follow the standard disruption process:

Identify a list of prioritized candidates for the disruption method.
- If there are pods that cannot be evicted on the node, Karpenter will ignore the node and try disrupting it later.
- If there are no disruptable nodes, continue to the next disruption method.
For each disruptable node, execute a scheduling simulation with the pods on the node to find if any replacement nodes are needed.
Add the karpenter.sh/disruption:NoSchedule taint to the node(s) to prevent pods from scheduling to it.
Pre-spin any replacement nodes needed as calculated in Step (2), and wait for them to become ready.
- If a replacement node fails to initialize, un-taint the node(s), and restart from Step (1), starting at the first disruption method again.
Delete the node(s) and wait for the Termination Controller to gracefully shutdown the node(s).
Once the Termination Controller terminates the node, go back to Step (1), starting at the first disruption method again.

Termination Controller

When a Karpenter node is deleted, the Karpenter finalizer will block deletion and the APIServer will set the DeletionTimestamp on the node, allowing Karpenter to gracefully shutdown the node, modeled after Kubernetes Graceful Node Shutdown. Karpenter’s graceful shutdown process will:

Add the karpenter.sh/disruption:NoSchedule taint to the node to prevent pods from scheduling to it.
Begin evicting the pods on the node with the Kubernetes Eviction API to respect PDBs, while ignoring all daemonset pods and static pods. Wait for the node to be fully drained before proceeding to Step (3).
- While waiting, if the underlying NodeClaim for the node no longer exists, remove the finalizer to allow the APIServer to delete the node, completing termination.
Terminate the NodeClaim in the Cloud Provider.
Remove the finalizer from the node to allow the APIServer to delete the node, completing termination.

Manual Methods

Node Deletion: You can use kubectl to manually remove a single Karpenter node or nodeclaim. Since each Karpenter node is owned by a NodeClaim, deleting either the node or the nodeclaim will cause cascade deletion of the other:

# Delete a specific nodeclaim
kubectl delete nodeclaim $NODECLAIM_NAME

# Delete a specific node
kubectl delete node $NODE_NAME

# Delete all nodeclaims
kubectl delete nodeclaims --all

# Delete all nodes owned by any nodepool
kubectl delete nodes -l karpenter.sh/nodepool

# Delete all nodeclaims owned by a specific nodepoolXS
kubectl delete nodeclaims -l karpenter.sh/nodepool=$NODEPOOL_NAME

NodePool Deletion: NodeClaims are owned by the NodePool through an owner reference that launched them. Karpenter will gracefully terminate nodes through cascading deletion when the owning NodePool is deleted.

Note

By adding the finalizer, Karpenter improves the default Kubernetes process of node deletion. When you run kubectl delete node on a node without a finalizer, the node is deleted without triggering the finalization logic. The instance will continue running in EC2, even though there is no longer a node object for it. The kubelet isn’t watching for its own existence, so if a node is deleted, the kubelet doesn’t terminate itself. All the pod objects get deleted by a garbage collection process later, because the pods’ node is gone.

Automated Methods

Expiration: Karpenter will mark nodes as expired and disrupt them after they have lived a set number of seconds, based on the NodePool’s spec.disruption.expireAfter value. You can use node expiry to periodically recycle nodes due to security concerns.
Consolidation: Karpenter works to actively reduce cluster cost by identifying when:
- Nodes can be removed because the node is empty
- Nodes can be removed as their workloads will run on other nodes in the cluster.
- Nodes can be replaced with cheaper variants due to a change in the workloads.
Drift: Karpenter will mark nodes as drifted and disrupt nodes that have drifted from their desired specification. See Drift to see which fields are considered.
Interruption: Karpenter will watch for upcoming interruption events that could affect your nodes (health events, spot interruption, etc.) and will taint, drain, and terminate the node(s) ahead of the event to reduce workload disruption.

Defaults

Disruption is configured through the NodePool’s disruption block by the consolidationPolicy, expireAfter and consolidateAfter fields. Karpenter will configure these fields with the following values by default if they are not set:

spec:
 disruption:
 consolidationPolicy: WhenUnderutilized
 expireAfter: 720h

Consolidation

Karpenter has two mechanisms for cluster consolidation:

Deletion - A node is eligible for deletion if all of its pods can run on free capacity of other nodes in the cluster.
Replace - A node can be replaced if all of its pods can run on a combination of free capacity of other nodes in the cluster and a single cheaper replacement node.

Consolidation has three mechanisms that are performed in order to attempt to identify a consolidation action:

Empty Node Consolidation - Delete any entirely empty nodes in parallel
Multi Node Consolidation - Try to delete two or more nodes in parallel, possibly launching a single replacement that is cheaper than the price of all nodes being removed
Single Node Consolidation - Try to delete any single node, possibly launching a single replacement that is cheaper than the price of that node

It’s impractical to examine all possible consolidation options for multi-node consolidation, so Karpenter uses a heuristic to identify a likely set of nodes that can be consolidated. For single-node consolidation we consider each node in the cluster individually.

When there are multiple nodes that could be potentially deleted or replaced, Karpenter chooses to consolidate the node that overall disrupts your workloads the least by preferring to terminate:

Nodes running fewer pods
Nodes that will expire soon
Nodes with lower priority pods

If consolidation is enabled, Karpenter periodically reports events against nodes that indicate why the node can’t be consolidated. These events can be used to investigate nodes that you expect to have been consolidated, but still remain in your cluster.

Events:
 Type Reason Age From Message
 ---- ------ ---- ---- -------
 Normal Unconsolidatable 66s karpenter pdb default/inflate-pdb prevents pod evictions
 Normal Unconsolidatable 33s (x3 over 30m) karpenter can't replace with a cheaper node

Warning

Using preferred anti-affinity and topology spreads can reduce the effectiveness of consolidation. At node launch, Karpenter attempts to satisfy affinity and topology spread preferences. In order to reduce node churn, consolidation must also attempt to satisfy these constraints to avoid immediately consolidating nodes after they launch. This means that consolidation may not disrupt nodes in order to avoid violating preferences, even if kube-scheduler can fit the host pods elsewhere. Karpenter reports these pods via logging to bring awareness to the possible issues they can cause (e.g. pod default/inflate-anti-self-55894c5d8b-522jd has a preferred Anti-Affinity which can prevent consolidation).

Note

For spot nodes, Karpenter only uses the deletion consolidation mechanism. It will not replace a spot node with a cheaper spot node. Spot instance types are selected with the price-capacity-optimized strategy and often the cheapest spot instance type is not launched due to the likelihood of interruption. Consolidation would then replace the spot instance with a cheaper instance negating the price-capacity-optimized strategy entirely and increasing interruption rate.

Drift

Drift handles changes to the NodePool/EC2NodeClass. For Drift, values in the NodePool/EC2NodeClass are reflected in the NodeClaimTemplateSpec/EC2NodeClassSpec in the same way that they’re set. A NodeClaim will be detected as drifted if the values in its owning NodePool/EC2NodeClass do not match the values in the NodeClaim. Similar to the upstream deployment.spec.template relationship to pods, Karpenter will annotate the owning NodePool and EC2NodeClass with a hash of the NodeClaimTemplateSpec to check for drift. Some special cases will be discovered either from Karpenter or through the CloudProvider interface, triggered by NodeClaim/Instance/NodePool/EC2NodeClass changes.

Special Cases on Drift

In special cases, drift can correspond to multiple values and must be handled differently. Drift on resolved fields can create cases where drift occurs without changes to CRDs, or where CRD changes do not result in drift. For example, if a NodeClaim has node.kubernetes.io/instance-type: m5.large, and requirements change from node.kubernetes.io/instance-type In [m5.large] to node.kubernetes.io/instance-type In [m5.large, m5.2xlarge], the NodeClaim will not be drifted because its value is still compatible with the new requirements. Conversely, if a NodeClaim is using a NodeClaim image ami: ami-abc, but a new image is published, Karpenter’s EC2NodeClass.spec.amiSelectorTerms will discover that the new correct value is ami: ami-xyz, and detect the NodeClaim as drifted.

NodePool

Fields
spec.template.spec.requirements

EC2NodeClass

Fields
spec.subnetSelectorTerms
spec.securityGroupSelectorTerms
spec.amiSelectorTerms

Behavioral Fields

Behavioral Fields are treated as over-arching settings on the NodePool to dictate how Karpenter behaves. These fields don’t correspond to settings on the NodeClaim or instance. They’re set by the user to control Karpenter’s Provisioning and disruption logic. Since these don’t map to a desired state of NodeClaims, behavioral fields are not considered for Drift.

NodePool

Fields
spec.weight
spec.limits
spec.disruption.*

Read the Drift Design for more.

To enable the drift feature flag, refer to the Feature Gates.

Karpenter will add the Drifted status condition on NodeClaims if the NodeClaim is drifted from its owning NodePool. Karpenter will also remove the Drifted status condition if either:

The Drift feature gate is not enabled but the NodeClaim is drifted, Karpenter will remove the status condition.
The NodeClaim isn’t drifted, but has the status condition, Karpenter will remove it.

Interruption

If interruption-handling is enabled, Karpenter will watch for upcoming involuntary interruption events that would cause disruption to your workloads. These interruption events include:

Spot Interruption Warnings
Scheduled Change Health Events (Maintenance Events)
Instance Terminating Events
Instance Stopping Events

When Karpenter detects one of these events will occur to your nodes, it automatically taints, drains, and terminates the node(s) ahead of the interruption event to give the maximum amount of time for workload cleanup prior to compute disruption. This enables scenarios where the terminationGracePeriod for your workloads may be long or cleanup for your workloads is critical, and you want enough time to be able to gracefully clean-up your pods.

For Spot interruptions, the NodePool will start a new node as soon as it sees the Spot interruption warning. Spot interruptions have a 2 minute notice before Amazon EC2 reclaims the instance. Once Karpenter has received this warning it will begin draining the node while in parallel provisioning a new node. Karpenter’s average node startup time means that, generally, there is sufficient time for the new node to become ready before EC2 initiates termination for the spot instance.

Note

Karpenter publishes Kubernetes events to the node for all events listed above in addition to Spot Rebalance Recommendations. Karpenter does not currently support taint, drain, and terminate logic for Spot Rebalance Recommendations.

If you require handling for Spot Rebalance Recommendations, you can use the AWS Node Termination Handler (NTH) alongside Karpenter; however, note that the AWS Node Termination Handler cordons and drains nodes on rebalance recommendations, potentially causing more node churn in the cluster than with interruptions alone. Further information can be found in the Troubleshooting Guide.

Karpenter enables this feature by watching an SQS queue which receives critical events from AWS services which may affect your nodes. Karpenter requires that an SQS queue be provisioned and EventBridge rules and targets be added that forward interruption events from AWS services to the SQS queue. Karpenter provides details for provisioning this infrastructure in the CloudFormation template in the Getting Started Guide.

To enable interruption handling, configure the --interruption-queue CLI argument with the name of the interruption queue provisioned to handle interruption events.

Controls

Pod-Level Controls

Pods with blocking PDBs will not be evicted by the Termination Controller or be considered for voluntary disruption actions. When multiple pods on a node have different PDBs, none of the PDBs may be blocking for Karpenter to voluntary disrupt a node. This can create complex eviction scenarios:

If a pod matches multiple PDBs (via label selectors), ALL of these PDBs must allow for disruption
When different pods on the same node belong to different PDBs, ALL PDBs must simultaneously permit eviction
A single blocking PDB can prevent the entire node from being voluntary disrupted

For example, consider a node with these pods and PDBs:

Pod A: Matches PDB-1 (maxUnavailable: 0) and PDB-2 (maxUnavailable: 1)
Pod B: Matches PDB-3 (minAvailable: 100%)
Pod C: No PDB

In this scenario, Karpenter cannot voluntary disrupt the node because:

Pod A is blocked by PDB-1 even though PDB-2 would allow disruption
Pod B is blocked by PDB-3’s requirement for 100% availability

As seen in this example, the more PDBs there are affecting a Node, the more difficult it will be for Karpenter to find an opportunity to perform voluntary disruption actions.

Secondly, you can block Karpenter from voluntarily choosing to disrupt certain pods by setting the karpenter.sh/do-not-disrupt: "true" annotation on the pod. This is useful for pods that you want to run from start to finish without disruption. By opting pods out of this disruption, you are telling Karpenter that it should not voluntarily remove a node containing this pod.

Examples of pods that you might want to opt-out of disruption include an interactive game that you don’t want to interrupt or a long batch job (such as you might have with machine learning) that would need to start over if it were interrupted.

apiVersion: apps/v1
kind: Deployment
spec:
 template:
 metadata:
 annotations:
 karpenter.sh/do-not-disrupt: "true"

Note

This annotation will be ignored for terminating pods, terminal pods (Failed/Succeeded), DaemonSet pods, or static pods.

Examples of voluntary node removal that will be prevented by this annotation include:

Consolidation
Drift
Expiration

Note

Voluntary node removal does not include Interruption or manual deletion initiated through kubectl delete node. Both of these are considered involuntary events, since node removal cannot be delayed.

Node-Level Controls

You can block Karpenter from voluntarily choosing to disrupt certain nodes by setting the karpenter.sh/do-not-disrupt: "true" annotation on the node. This will prevent disruption actions on the node.

apiVersion: v1
kind: Node
metadata:
 annotations:
 karpenter.sh/do-not-disrupt: "true"

Example: Disable Disruption on a NodePool

NodePool .spec.annotations allow you to set annotations that will be applied to all nodes launched by this NodePool. By setting the annotation karpenter.sh/do-not-disrupt: "true" on the NodePool, you will selectively prevent all nodes launched by this NodePool from being considered in disruption actions.

apiVersion: karpenter.sh/v1beta1
kind: NodePool
metadata:
 name: default
spec:
 template:
 metadata:
 annotations: # will be applied to all nodes
 karpenter.sh/do-not-disrupt: "true"

Karpenter – Concepts

V0.32: NodePools

spec.template.spec.requirements

Instance Types

Availability Zones

Architecture

Operating System

Capacity Type

Recommended

Note

spec.template.spec.nodeClassRef

spec.template.spec.kubelet

Reserved Resources

Note

Eviction Thresholds

Supported Eviction Signals

Soft Eviction Grace Periods

Pod Density

Note

Windows Support Notice

Max Pods

Note

Windows Support Notice

Pods Per Core

Note

Pods Per Core on Bottlerocket

spec.disruption

spec.limits

Note

spec.weight

Examples

Isolating Expensive Hardware

Cilium Startup Taint

V0.32: NodeClasses

spec.amiFamily

AL2

Bottlerocket

Ubuntu

Windows2019

Windows2022

Note

Custom

spec.subnetSelectorTerms

Tip

Examples

spec.securityGroupSelectorTerms

Tip

Note

Examples

spec.amiSelectorTerms

Note

Tip

Note

Examples

spec.role

spec.instanceProfile

Note

spec.tags

Note

spec.metadataOptions

spec.blockDeviceMappings

AL2

Bottlerocket

Ubuntu

Windows2019/Windows2022

Custom

spec.userData

Note

AL2/Ubuntu

Passed-in UserData (bash)

Passed-in UserData (MIME)

Merged UserData

Note

Bottlerocket

Passed-in UserData

Merged UserData

Windows2019/Windows2022

Passed-in UserData

Merged UserData

Windows Support Notice

`Exists` Operator