Cannot create NSX manager nodes from the UI after single-node restoration

Products

• VMware NSX

Issue/Introduction

• After restoring a single node of an NSX cluster, you are unable to create additional NSX manager nodes from the UI.
• Navigate to System > Fabric > Compute Managers - The vCenter Server connection remains stuck in a "connecting" state.
  
• The restore process status shows as "Suspended For Manual Steps".
  
• You may observe the following status in the /var/log/cm-inventory/cm-inventory.log file:

  <Timestamp> INFO ClusterNodeConfigListener-1-1 CmInventoryService 115193 FABRIC [nsx@6876 comp="nsx-manager" level="INFO" subcomp="cm-inventory"] CM status: >connection_status: COMPUTE_MANAGER_CONNECTION_STATUS_CONNECTING

Environment

VMware NSX
VMware vCenter Server

Cause

The NSX manager connection workflow will not allow the compute manager to be changed or updated until the cluster restore process is fully complete.

Resolution

To proceed, you can use one of the following two workarounds. Choose the option that best suits your environment size and requirements:

• Workaround 1: For small and medium Environments, proceed with the restore on a Single Manager.
  1.  Acknowledge the suspension and proceed with completing the restore operation on the single NSX manager node.
     
  2. Wait for the restore process to fully complete.
  3. Wait for the compute manager to get into a connected state.
  4. Deploy additional nodes from the UI.
      
• Workaround 2: For large Environments with 100+ hosts, Manually Deploy Additional Managers.
  1. Do not use the automated UI option to deploy the additional nodes.
  2. Manually deploy two additional NSX manager appliances on the same vCenter Server.
  3. Manually join the newly deployed manager nodes to the first manager node to form the cluster. Refer to Form an NSX Manager Cluster Using the CLI
  4. Resume the restoration.