Pause machinedeployment on a workload cluster to perform a configuration change on the underlying VM

Products

• VMware Telco Cloud Automation
• VMware Telco Cloud Platform

Issue/Introduction

• Powered off VMs that make up a workload cluster keep restarting
• Unable to power off associated VMs

Environment

TCA: 3.2

TCP: 5.0

Cause

The workload cluster's lifecycle management, specifically the MachineDeployment controller, is enforcing the desired state of the nodes.

Resolution

1. SSH into the associated management cluster VIP using the capv user

2. Obtain machinedeployment name and namespace

   kubectl get machinedeployment -A

3. Pause the sopecifed machinedeployment

   kubectl annotate machinedeployment <deployment-name> -n <namespace> cluster.x-k8s.io/paused=""

4. Confirm the annotation

   Search for : metadata.annotations.cluster.x-k8s.io/paused: ""

5. You can now power off the associated VMs in vCenter. The VMs will remain off.

6. Revert changes

   kubectl annotate machinedeployment <deployment-name> -n <namespace>  cluster.x-k8s.io/paused-