In a Tanzu Kubernetes cluster, the Supervisor cluster continuously monitors and reconciles the cluster. If any worker node fails to respond within the monitoring interval, it is marked as failed, and a new worker node is created to replace it. However, this operation may not always be accurate. If you plan to perform any maintenance on the Tanzu environment, it is necessary to pause the cluster reconciliation; otherwise, you may see many nodes being recreated.
If you are planning any network maintenance within your organization and expect any storage latency or network latency/interruption, it is better to pause the cluster reconciliation.
How to Pause the Cluster Reconciliation
Login to the Supervisor Node:
root@vcenter [ ~ ]# sshpass -p /usr/lib/vmware-wcp/decryptK8Pwd.py |grep PWD |awk '{print $2}' ssh -o StrictHostKeyChecking=no root@/usr/lib/vmware-wcp/decryptK8Pwd.py |grep IP |awk '{print $2}'
To Pause the Cluster Reconciliation:
kubectl patch cluster --type merge -p '{"spec":{"paused": true}}' -n tanzu-vns tanzu-cluster
To Resume the Cluster Reconciliation:
kubectl patch cluster --type merge -p '{"spec":{"paused": false}}' -n tanzu-vns tanzu-cluster
Top comments (0)