Manual steps for adding a vRA/vRO 8.0 node into a vRA/vRO 8.0 cluster when another node has been previously removed

Products

• VMware Aria Suite

Issue/Introduction

Symptoms:
Adding a new node fails during the kubeadm step when running vracli cluster join

Environment

VMware vRealize Orchestrator 8.x
VMware vRealize Automation 8.x

Cause

A service endpoint remains for previously-removed nodes. When adding new nodes this endpoint is contacted and as its is not available adding a new node fails.

Resolution

Steps to remove it:
 

1. SSH login to one of the nodes in the cluster
2. Run kubectl edit configmaps -n kube-system kubeadm-config
3. To remove the bad node endpoint locate its configuration and delete the following lines:
 

ClusterStatus: |
apiEndpoints:
sc1-10-78-211-219.eng.vmware.com:
advertiseAddress: 10.78.211.219
bindPort: 6443
sc1-10-78-212-83.eng.vmware.com:
advertiseAddress: 10.78.212.83
bindPort: 6443
sc1-10-78-212-116.eng.vmware.com:
advertiseAddress: 10.78.212.116
bindPort: 6443
 

For example to clean up node sc1-10-78-212-116.eng.vmware.com, delete:
 

sc1-10-78-212-116.eng.vmware.com:
advertiseAddress: 10.78.212.116
bindPort: 6443
 

4. Retry adding the new node to the cluster.