Amazon EKS — Common Errors Reference

Common EKS Errors Reference

Quick lookup for the most common EKS error messages and their solutions.

---

Pod Errors

Error	Cause	Fix
CrashLoopBackOff	App crashes on start	Check kubectl logs --previous
ImagePullBackOff	Can’t pull container image	Check image name, ECR auth, node IAM
ErrImagePull	Image doesn’t exist	Verify image:tag in registry
OOMKilled	Container exceeded memory limit	Increase resources.limits.memory
Evicted	Node under resource pressure	Check node conditions, add capacity
CreateContainerConfigError	Missing ConfigMap or Secret	Verify referenced resources exist
RunContainerError	Container runtime failure	Check security context, capabilities

---

Node Errors

Error	Cause	Fix
NotReady	Kubelet or networking issue	Check instance health, kubelet logs
MemoryPressure	Node memory exhausted	Evict pods or add nodes
DiskPressure	Node disk full	Clean images: docker system prune
Taint NoSchedule	Node manually tainted	Remove taint or add toleration

---

Networking Errors

Error	Cause	Fix
dial tcp: lookup ... no such host	DNS resolution failure	Check CoreDNS pods
connection refused	Target port not open	Verify service ports and endpoints
connection timed out	Security group/NP blocking	Check SGs and NetworkPolicies
pod CIDR exhausted	Too many pods for subnet	Use custom networking or larger subnet

---

Storage Errors

Error	Cause	Fix
PVC Pending	No matching PV or StorageClass	Check SC exists and CSI driver running
FailedAttachVolume	EBS in wrong AZ	Use WaitForFirstConsumer
FailedMount	Mount target issue	Check EFS mount targets in subnet
permission denied	Wrong UID/GID	Set securityContext.fsGroup

---

IAM / Auth Errors

Error	Cause	Fix
Unauthorized	User not in aws-auth	Add IAM role to aws-auth ConfigMap
AccessDenied	Missing IAM permission	Check IRSA policy
ExpiredTokenException	AWS credentials expired	Refresh kubeconfig: aws eks update-kubeconfig
could not get token	AWS CLI not configured	Run aws configure

---

Quick Debug Commands

# Overall cluster health
$ kubectl get nodes
$ kubectl get pods -A | grep -v Running

# Events (newest first)
$ kubectl get events --sort-by='.lastTimestamp' -A | tail -20

# Resource usage
$ kubectl top nodes
$ kubectl top pods -A --sort-by=cpu | head -10

# API server connectivity
$ kubectl cluster-info

# Check all system components
$ kubectl get pods -n kube-system

---

What’s Next?

Practice: Troubleshooting Exercises.