Migrating from Cluster Autoscaler

Migrate to Karpenter from Cluster Autoscaler

This guide will show you how to switch from the Kubernetes Cluster Autoscaler to Karpenter for automatic node provisioning. We will make the following assumptions in this guide

  • You will use an existing EKS cluster
  • You will use existing VPC and subnets
  • You will use existing security groups
  • Your nodes are part of one or more node groups
  • Your workloads have pod disruption budgets that adhere to EKS best practices
  • Your cluster has an OIDC provider for service accounts

This guide will also assume you have the aws CLI installed. You can also perform many of these steps in the console, but we will use the command line for simplicity.

Set a variable for your cluster name.

KARPENTER_NAMESPACE=kube-system
CLUSTER_NAME=<your cluster name>

Set other variables from your cluster configuration.

AWS_PARTITION="aws" # if you are not using standard partitions, you may need to configure to aws-cn / aws-us-gov
AWS_REGION="$(aws configure list | grep region | tr -s " " | cut -d" " -f3)"
OIDC_ENDPOINT="$(aws eks describe-cluster --name ${CLUSTER_NAME} \
    --query "cluster.identity.oidc.issuer" --output text)"
AWS_ACCOUNT_ID=$(aws sts get-caller-identity --query 'Account' \
    --output text)

Use that information to create our IAM roles, inline policy, and trust relationship.

Create IAM roles

To get started with our migration we first need to create two new IAM roles for nodes provisioned with Karpenter and the Karpenter controller.

To create the Karpenter node role we will use the following policy and commands.

echo '{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Effect": "Allow",
            "Principal": {
                "Service": "ec2.amazonaws.com"
            },
            "Action": "sts:AssumeRole"
        }
    ]
}' > node-trust-policy.json

aws iam create-role --role-name "KarpenterNodeRole-${CLUSTER_NAME}" \
    --assume-role-policy-document file://node-trust-policy.json

Now attach the required policies to the role

aws iam attach-role-policy --role-name "KarpenterNodeRole-${CLUSTER_NAME}" \
    --policy-arn arn:${AWS_PARTITION}:iam::aws:policy/AmazonEKSWorkerNodePolicy

aws iam attach-role-policy --role-name "KarpenterNodeRole-${CLUSTER_NAME}" \
    --policy-arn arn:${AWS_PARTITION}:iam::aws:policy/AmazonEKS_CNI_Policy

aws iam attach-role-policy --role-name "KarpenterNodeRole-${CLUSTER_NAME}" \
    --policy-arn arn:${AWS_PARTITION}:iam::aws:policy/AmazonEC2ContainerRegistryReadOnly

aws iam attach-role-policy --role-name "KarpenterNodeRole-${CLUSTER_NAME}" \
    --policy-arn arn:${AWS_PARTITION}:iam::aws:policy/AmazonSSMManagedInstanceCore

Now we need to create an IAM role that the Karpenter controller will use to provision new instances. The controller will be using IAM Roles for Service Accounts (IRSA) which requires an OIDC endpoint.

If you have another option for using IAM credentials with workloads (e.g. kube2iam) your steps will be different.

cat << EOF > controller-trust-policy.json
{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Effect": "Allow",
            "Principal": {
                "Federated": "arn:${AWS_PARTITION}:iam::${AWS_ACCOUNT_ID}:oidc-provider/${OIDC_ENDPOINT#*//}"
            },
            "Action": "sts:AssumeRoleWithWebIdentity",
            "Condition": {
                "StringEquals": {
                    "${OIDC_ENDPOINT#*//}:aud": "sts.amazonaws.com",
                    "${OIDC_ENDPOINT#*//}:sub": "system:serviceaccount:${KARPENTER_NAMESPACE}:karpenter"
                }
            }
        }
    ]
}
EOF

aws iam create-role --role-name KarpenterControllerRole-${CLUSTER_NAME} \
    --assume-role-policy-document file://controller-trust-policy.json

cat << EOF > controller-policy.json
{
    "Statement": [
        {
            "Action": [
                "ssm:GetParameter",
                "ec2:DescribeImages",
                "ec2:RunInstances",
                "ec2:DescribeSubnets",
                "ec2:DescribeSecurityGroups",
                "ec2:DescribeLaunchTemplates",
                "ec2:DescribeInstances",
                "ec2:DescribeInstanceTypes",
                "ec2:DescribeInstanceTypeOfferings",
                "ec2:DescribeAvailabilityZones",
                "ec2:DeleteLaunchTemplate",
                "ec2:CreateTags",
                "ec2:CreateLaunchTemplate",
                "ec2:CreateFleet",
                "ec2:DescribeSpotPriceHistory",
                "pricing:GetProducts"
            ],
            "Effect": "Allow",
            "Resource": "*",
            "Sid": "Karpenter"
        },
        {
            "Action": "ec2:TerminateInstances",
            "Condition": {
                "StringLike": {
                    "ec2:ResourceTag/karpenter.sh/nodepool": "*"
                }
            },
            "Effect": "Allow",
            "Resource": "*",
            "Sid": "ConditionalEC2Termination"
        },
        {
            "Effect": "Allow",
            "Action": "iam:PassRole",
            "Resource": "arn:${AWS_PARTITION}:iam::${AWS_ACCOUNT_ID}:role/KarpenterNodeRole-${CLUSTER_NAME}",
            "Sid": "PassNodeIAMRole"
        },
        {
            "Effect": "Allow",
            "Action": "eks:DescribeCluster",
            "Resource": "arn:${AWS_PARTITION}:eks:${AWS_REGION}:${AWS_ACCOUNT_ID}:cluster/${CLUSTER_NAME}",
            "Sid": "EKSClusterEndpointLookup"
        },
        {
            "Sid": "AllowScopedInstanceProfileCreationActions",
            "Effect": "Allow",
            "Resource": "*",
            "Action": [
            "iam:CreateInstanceProfile"
            ],
            "Condition": {
            "StringEquals": {
                "aws:RequestTag/kubernetes.io/cluster/${CLUSTER_NAME}": "owned",
                "aws:RequestTag/topology.kubernetes.io/region": "${AWS_REGION}"
            },
            "StringLike": {
                "aws:RequestTag/karpenter.k8s.aws/ec2nodeclass": "*"
            }
            }
        },
        {
            "Sid": "AllowScopedInstanceProfileTagActions",
            "Effect": "Allow",
            "Resource": "*",
            "Action": [
            "iam:TagInstanceProfile"
            ],
            "Condition": {
            "StringEquals": {
                "aws:ResourceTag/kubernetes.io/cluster/${CLUSTER_NAME}": "owned",
                "aws:ResourceTag/topology.kubernetes.io/region": "${AWS_REGION}",
                "aws:RequestTag/kubernetes.io/cluster/${CLUSTER_NAME}": "owned",
                "aws:RequestTag/topology.kubernetes.io/region": "${AWS_REGION}"
            },
            "StringLike": {
                "aws:ResourceTag/karpenter.k8s.aws/ec2nodeclass": "*",
                "aws:RequestTag/karpenter.k8s.aws/ec2nodeclass": "*"
            }
            }
        },
        {
            "Sid": "AllowScopedInstanceProfileActions",
            "Effect": "Allow",
            "Resource": "*",
            "Action": [
            "iam:AddRoleToInstanceProfile",
            "iam:RemoveRoleFromInstanceProfile",
            "iam:DeleteInstanceProfile"
            ],
            "Condition": {
            "StringEquals": {
                "aws:ResourceTag/kubernetes.io/cluster/${CLUSTER_NAME}": "owned",
                "aws:ResourceTag/topology.kubernetes.io/region": "${AWS_REGION}"
            },
            "StringLike": {
                "aws:ResourceTag/karpenter.k8s.aws/ec2nodeclass": "*"
            }
            }
        },
        {
            "Sid": "AllowInstanceProfileReadActions",
            "Effect": "Allow",
            "Resource": "*",
            "Action": "iam:GetInstanceProfile"
        }
    ],
    "Version": "2012-10-17"
}
EOF

aws iam put-role-policy --role-name KarpenterControllerRole-${CLUSTER_NAME} \
    --policy-name KarpenterControllerPolicy-${CLUSTER_NAME} \
    --policy-document file://controller-policy.json

Add tags to subnets and security groups

We need to add tags to our nodegroup subnets so Karpenter will know which subnets to use.

for NODEGROUP in $(aws eks list-nodegroups --cluster-name ${CLUSTER_NAME} \
    --query 'nodegroups' --output text); do aws ec2 create-tags \
        --tags "Key=karpenter.sh/discovery,Value=${CLUSTER_NAME}" \
        --resources $(aws eks describe-nodegroup --cluster-name ${CLUSTER_NAME} \
        --nodegroup-name $NODEGROUP --query 'nodegroup.subnets' --output text )
done

Add tags to our security groups. This command only tags the security groups for the first nodegroup in the cluster. If you have multiple nodegroups or multiple security groups you will need to decide which one Karpenter should use.

NODEGROUP=$(aws eks list-nodegroups --cluster-name ${CLUSTER_NAME} \
    --query 'nodegroups[0]' --output text)

LAUNCH_TEMPLATE=$(aws eks describe-nodegroup --cluster-name ${CLUSTER_NAME} \
    --nodegroup-name ${NODEGROUP} --query 'nodegroup.launchTemplate.{id:id,version:version}' \
    --output text | tr -s "\t" ",")

# If your EKS setup is configured to use only Cluster security group, then please execute -

SECURITY_GROUPS=$(aws eks describe-cluster \
    --name ${CLUSTER_NAME} --query "cluster.resourcesVpcConfig.clusterSecurityGroupId" --output text)

# If your setup uses the security groups in the Launch template of a managed node group, then :

SECURITY_GROUPS=$(aws ec2 describe-launch-template-versions \
    --launch-template-id ${LAUNCH_TEMPLATE%,*} --versions ${LAUNCH_TEMPLATE#*,} \
    --query 'LaunchTemplateVersions[0].LaunchTemplateData.[NetworkInterfaces[0].Groups||SecurityGroupIds]' \
    --output text)

aws ec2 create-tags \
    --tags "Key=karpenter.sh/discovery,Value=${CLUSTER_NAME}" \
    --resources ${SECURITY_GROUPS}

Update aws-auth ConfigMap

We need to allow nodes that are using the node IAM role we just created to join the cluster. To do that we have to modify the aws-auth ConfigMap in the cluster.

kubectl edit configmap aws-auth -n kube-system

You will need to add a section to the mapRoles that looks something like this. Replace the ${AWS_PARTITION} variable with the account partition, ${AWS_ACCOUNT_ID} variable with your account ID, and ${CLUSTER_NAME} variable with the cluster name, but do not replace the {{EC2PrivateDNSName}}.

- groups:
  - system:bootstrappers
  - system:nodes
  ## If you intend to run Windows workloads, the kube-proxy group should be specified.
  # For more information, see https://github.com/aws/karpenter/issues/5099.
  # - eks:kube-proxy-windows
  rolearn: arn:${AWS_PARTITION}:iam::${AWS_ACCOUNT_ID}:role/KarpenterNodeRole-${CLUSTER_NAME}
  username: system:node:{{EC2PrivateDNSName}}

The full aws-auth configmap should have two groups. One for your Karpenter node role and one for your existing node group.

Deploy Karpenter

First set the Karpenter release you want to deploy.

export KARPENTER_VERSION=v0.34.1

We can now generate a full Karpenter deployment yaml from the helm chart.

helm template karpenter oci://public.ecr.aws/karpenter/karpenter --version "${KARPENTER_VERSION}" --namespace "${KARPENTER_NAMESPACE}" \
    --set "settings.clusterName=${CLUSTER_NAME}" \
    --set "serviceAccount.annotations.eks\.amazonaws\.com/role-arn=arn:${AWS_PARTITION}:iam::${AWS_ACCOUNT_ID}:role/KarpenterControllerRole-${CLUSTER_NAME}" \
    --set controller.resources.requests.cpu=1 \
    --set controller.resources.requests.memory=1Gi \
    --set controller.resources.limits.cpu=1 \
    --set controller.resources.limits.memory=1Gi > karpenter.yaml

Modify the following lines in the karpenter.yaml file.

Set node affinity

Edit the karpenter.yaml file and find the karpenter deployment affinity rules. Modify the affinity so karpenter will run on one of the existing node group nodes.

The rules should look something like this. Modify the value to match your $NODEGROUP, one node group per line.

affinity:
  nodeAffinity:
    requiredDuringSchedulingIgnoredDuringExecution:
      nodeSelectorTerms:
      - matchExpressions:
        - key: karpenter.sh/nodepool
          operator: DoesNotExist
      - matchExpressions:
        - key: eks.amazonaws.com/nodegroup
          operator: In
          values:
          - ${NODEGROUP}
  podAntiAffinity:
    requiredDuringSchedulingIgnoredDuringExecution:
      - topologyKey: "kubernetes.io/hostname"

Now that our deployment is ready we can create the karpenter namespace, create the NodePool CRD, and then deploy the rest of the karpenter resources.

kubectl create namespace "${KARPENTER_NAMESPACE}" || true
kubectl create -f \
    https://raw.githubusercontent.com/aws/karpenter-provider-aws/${KARPENTER_VERSION}/pkg/apis/crds/karpenter.sh_nodepools.yaml
kubectl create -f \
    https://raw.githubusercontent.com/aws/karpenter-provider-aws/${KARPENTER_VERSION}/pkg/apis/crds/karpenter.k8s.aws_ec2nodeclasses.yaml
kubectl create -f \
    https://raw.githubusercontent.com/aws/karpenter-provider-aws/${KARPENTER_VERSION}/pkg/apis/crds/karpenter.sh_nodeclaims.yaml
kubectl apply -f karpenter.yaml

Create default NodePool

We need to create a default NodePool so Karpenter knows what types of nodes we want for unscheduled workloads. You can refer to some of the example NodePool for specific needs.

cat <<EOF | envsubst | kubectl apply -f -
apiVersion: karpenter.sh/v1beta1
kind: NodePool
metadata:
  name: default
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: ["spot"]
        - key: karpenter.k8s.aws/instance-category
          operator: In
          values: ["c", "m", "r"]
        - key: karpenter.k8s.aws/instance-generation
          operator: Gt
          values: ["2"]
      nodeClassRef:
        name: default
  limits:
    cpu: 1000
  disruption:
    consolidationPolicy: WhenUnderutilized
    expireAfter: 720h # 30 * 24h = 720h
---
apiVersion: karpenter.k8s.aws/v1beta1
kind: EC2NodeClass
metadata:
  name: default
spec:
  amiFamily: AL2 # Amazon Linux 2
  role: "KarpenterNodeRole-${CLUSTER_NAME}" # replace with your cluster name
  subnetSelectorTerms:
    - tags:
        karpenter.sh/discovery: "${CLUSTER_NAME}" # replace with your cluster name
  securityGroupSelectorTerms:
    - tags:
        karpenter.sh/discovery: "${CLUSTER_NAME}" # replace with your cluster name
EOF

Set nodeAffinity for critical workloads (optional)

You may also want to set a nodeAffinity for other critical cluster workloads.

Some examples are

  • coredns
  • metric-server

You can edit them with kubectl edit deploy ... and you should add node affinity for your static node group instances. Modify the value to match your $NODEGROUP, one node group per line.

affinity:
  nodeAffinity:
    requiredDuringSchedulingIgnoredDuringExecution:
      nodeSelectorTerms:
      - matchExpressions:
        - key: eks.amazonaws.com/nodegroup
          operator: In
          values:
          - ${NODEGROUP}

Remove CAS

Now that karpenter is running we can disable the cluster autoscaler. To do that we will scale the number of replicas to zero.

kubectl scale deploy/cluster-autoscaler -n kube-system --replicas=0

To get rid of the instances that were added from the node group we can scale our nodegroup down to a minimum size to support Karpenter and other critical services.

Note: If your workloads do not have pod disruption budgets set, the following command will cause workloads to be unavailable.

If you have a single multi-AZ node group, we suggest a minimum of 2 instances.

aws eks update-nodegroup-config --cluster-name ${CLUSTER_NAME} \
    --nodegroup-name ${NODEGROUP} \
    --scaling-config "minSize=2,maxSize=2,desiredSize=2"

Or, if you have multiple single-AZ node groups, we suggest a minimum of 1 instance each.

for NODEGROUP in $(aws eks list-nodegroups --cluster-name ${CLUSTER_NAME} \
    --query 'nodegroups' --output text); do aws eks update-nodegroup-config --cluster-name ${CLUSTER_NAME} \
    --nodegroup-name ${NODEGROUP} \
    --scaling-config "minSize=1,maxSize=1,desiredSize=1"
done

Verify Karpenter

As nodegroup nodes are drained you can verify that Karpenter is creating nodes for your workloads.

kubectl logs -f -n karpenter -c controller -l app.kubernetes.io/name=karpenter

You should also see new nodes created in your cluster as the old nodes are removed

kubectl get nodes
Last modified February 19, 2024: Update docs to point to v0.34.1 (293f6476)