What would you do if your EKS worker nodes keep spinning up but never join the cluster?
I’ve been there. Here how I would troubleshoot it:
1️⃣ Check IAM Role for worker nodes
– Go to IAM → Roles → AmazonEKSWorkerNodeRole (or custom role).
– Ensure it has policy attached. If you are missing permissions, your nodes won’t be able to connect.
2️⃣ Review Security Groups
– Confirm that the worker node security group allows inbound traffic from the EKS control plane (port 443).
– Also, ensure outbound traffic is not restricted, especially to cluster endpoints.
3️⃣ Check Node Registration with kubectl
– Run kubectl get nodes , to verify if nodes are missing
– Run kubectl get cs , to verify verify control plane health.
4️⃣ Restart the worker nodes and check logs for errors
– Sometimes a simple restart helps, but watch for kubelet errors.
What other troubleshooting techniques have you used?
Wanna be good at AWS? Learn to build, but also learn to fix things when they break.
Leave a Reply