nodeconfig-operator pod fails to schedule

Products

• VMware Telco Cloud Automation

Issue/Introduction

• The nodeconfig-operator pod is stuck in a 'Pending' state
• Describing the nodeconfig-operator pod shows in the 'Events' section that no nodes match the pod's selector

 

Environment

2.3

Cause

Kubernetes versions 1.23 and earlier used the label 'master' to identify control-plane nodes, where versions 1.24 and later no longer use this label. The nodeconfig-operator pod is still configured with the legacy 'master' nodeSelector.

Resolution

Remove the 'master' selectorTerm from the nodeconfig-operator deployment:

1. kubectl edit deployment -n tca-system nodeconfig-operator

spec:
  affinity:
    nodeAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:
        nodeSelectorTerms:
        - matchExpressions:
          - key: node-role.kubernetes.io/control-plane
            operator: Exists
 
2. Confirm the nodeconfig-operator pod is scheduled to one of the control-plane nodes and enters a Running state:
   
   kubectl get pods -n tca-system -o wide | grep nodeconfig-operator