Kubernetes Pod Stuck in Container Creating State Due to Node Freezing

Products

• VMware Telco Cloud Automation
• VMware Telco Cloud Platform

Issue/Introduction

• The pod becomes stuck in a "container creating" state.
• The node assigned to the stuck pod is inaccessible via SSH, and the VM console is unresponsive.
• The vCenter monitor page shows the node VM CPU usage pegged at 100%.

Environment

TCA 3.4

TCP 5.1

Cause

The underlying Kubernetes node VM became unresponsive/frozen due to a 100% CPU utilization spike, which caused the local kubelet to stall during the container creation phase.

Resolution

1. Identify the unresponsive node VM that is assigned to the stuck pod.
   • Run the following command from the cluster's control plane:
     • kubectl get pod -n <namespace> <podName> -o wide
2. Delete the unresponsive node VM from vCenter.
3. Allow the Kubernetes cluster machine controllers to automatically provision and recreate a new, healthy node.
4. Verify that pending pods successfully schedule, initialize, and transition out of the "container creating" state.