Kubernets cluster pod stays in pending state

Question

I've created a kubernetes cluster on my linux machine. The problem is one of my pod is still in the pending stage. Is there a way to troubleshoot this?

Answer 1

If a Pod is stuck in Pending it means that it can not be scheduled onto a node. Generally this is because there are insufficient resources of one type or another that prevent scheduling. Look at the output of the kubectl describe ... command above. There should be messages from the scheduler about why it can not schedule your pod. Reasons include:

• You don’t have enough resources: You may have exhausted the supply of CPU or Memory in your cluster, in this case you need to delete Pods, adjust resource requests, or add new nodes to your cluster. See Compute Resources document for more information.

• You are using hostPort: When you bind a Pod to a hostPort there are a limited number of places that pod can be scheduled. In most cases, hostPort is unnecessary, try using a Service object to expose your Pod. If you do require hostPort then you can only schedule as many Pods as there are nodes in your Kubernetes cluster.

Answer 2

Have a look at the Kubernetes cluster troubleshoot guide: https://kubernetes.io/docs/tasks/debug-application-cluster/debug-application/

Answer 3

This issue strikes when the pod is unable to connect to the master server. This is very common when you set up an EKS cluster. These are two things you need to keep in mind for the cluster to be fine.

• Enable private endpoint access so that the worker node or pod under VPC is able to connect
• Setup security group and map with EKS cluster setup. Do not forget to add worker security group in the ingress rule with 443 port access.

Answer 4

Hey, I had the exact same issue. I don't know what the problem was but I reset the cluster and created it again and it worked fine for me. If somebody knows the reason, please do let me know! Thanks

Answer 5

I had the same issue. I spent days trying to figure out the solution. So the problem was there was a mismatch with the hostname in master and the clients. Very ashamed of this stupidity but such mistakes can give you very annoying errors.