AWS EKS Cluster Autoscaler Not Scaling — Every Fix (2026)

kubectl get pods -n kube-system | grep cluster-autoscaler
 
# Check CA logs
kubectl logs -n kube-system \
  -l app.kubernetes.io/name=cluster-autoscaler \
  --tail=100

# Check pod status
kubectl describe pod <pending-pod>
 
# Look for Events section
Events:
  Warning  FailedScheduling  0/3 nodes available: 
    3 Insufficient cpu.   ← CA will scale up for this
 
  Warning  FailedScheduling  pod has unbound PVC  
    ← CA will NOT scale up for this (storage issue, not resources)

# Get your node group ASG name
aws eks describe-nodegroup \
  --cluster-name my-cluster \
  --nodegroup-name my-nodegroup \
  --query "nodegroup.resources.autoScalingGroups"
 
# Check tags on the ASG
aws autoscaling describe-auto-scaling-groups \
  --auto-scaling-group-names my-cluster-nodegroup-xxx \
  --query "AutoScalingGroups[].Tags"

aws autoscaling create-or-update-tags \
  --tags \
    ResourceId=my-asg-name,ResourceType=auto-scaling-group,Key=k8s.io/cluster-autoscaler/my-cluster,Value=owned,PropagateAtLaunch=false \
    ResourceId=my-asg-name,ResourceType=auto-scaling-group,Key=k8s.io/cluster-autoscaler/enabled,Value=true,PropagateAtLaunch=false

# Check CA's service account has IAM role
kubectl describe serviceaccount cluster-autoscaler -n kube-system
 
# Look for:
# Annotations:  eks.amazonaws.com/role-arn: arn:aws:iam::...

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Action": [
        "autoscaling:DescribeAutoScalingGroups",
        "autoscaling:DescribeAutoScalingInstances",
        "autoscaling:DescribeLaunchConfigurations",
        "autoscaling:DescribeScalingActivities",
        "autoscaling:DescribeTags",
        "autoscaling:SetDesiredCapacity",
        "autoscaling:TerminateInstanceInAutoScalingGroup",
        "ec2:DescribeImages",
        "ec2:DescribeInstanceTypes",
        "ec2:DescribeLaunchTemplateVersions",
        "ec2:GetInstanceTypesFromInstanceRequirements",
        "eks:DescribeNodegroup"
      ],
      "Resource": "*"
    }
  ]
}

# Create IAM role with IRSA
eksctl create iamserviceaccount \
  --cluster my-cluster \
  --namespace kube-system \
  --name cluster-autoscaler \
  --attach-policy-arn arn:aws:iam::123456789:policy/ClusterAutoscalerPolicy \
  --override-existing-serviceaccounts \
  --approve

# Check k8s version
kubectl version --short
 
# Check CA version
kubectl describe deployment cluster-autoscaler -n kube-system | grep Image
 
# CA 1.29 for k8s 1.29, CA 1.30 for k8s 1.30, etc.

# Replace with correct version
kubectl set image deployment/cluster-autoscaler \
  cluster-autoscaler=registry.k8s.io/autoscaling/cluster-autoscaler:v1.30.3 \
  -n kube-system

# Check what labels new nodes get
kubectl get node -l eks.amazonaws.com/nodegroup=my-nodegroup --show-labels
 
# Check what your pod requires
kubectl describe pod <pending-pod> | grep -A5 "Node-Selectors"

# In your EKS node group launch template user data:
--kubelet-extra-args '--node-labels=node-type=gpu'

# Check node taints
kubectl describe node <new-node> | grep Taints
 
# Check pod tolerations
kubectl describe pod <pending-pod> | grep -A5 Tolerations

# Check PDBs
kubectl get pdb -A
 
# Check if any PDB is blocking drain
kubectl describe pdb my-pdb

metadata:
  annotations:
    cluster-autoscaler.kubernetes.io/safe-to-evict: "false"

# Find pods with safe-to-evict=false
kubectl get pods -A -o json | \
  python3 -c "
import sys, json
data = json.load(sys.stdin)
for item in data['items']:
    ann = item['metadata'].get('annotations', {})
    if ann.get('cluster-autoscaler.kubernetes.io/safe-to-evict') == 'false':
        print(item['metadata']['namespace'], item['metadata']['name'])
"

#!/bin/bash
echo "=== CA Pod Status ==="
kubectl get pods -n kube-system -l app.kubernetes.io/name=cluster-autoscaler
 
echo ""
echo "=== Pending Pods ==="
kubectl get pods -A --field-selector=status.phase=Pending
 
echo ""
echo "=== Node Capacity ==="
kubectl describe nodes | grep -A4 "Allocated resources"
 
echo ""
echo "=== CA Recent Logs ==="
kubectl logs -n kube-system \
  -l app.kubernetes.io/name=cluster-autoscaler \
  --tail=30 | grep -E "scale|Scale|ERROR|WARN"