NSX Upgrade from 4.1.x to 4.2.x fails with "Failed to send HostConfig RPC" on Standalone Transport Nodes previously removed from vCenter

Products

• VMware NSX

Issue/Introduction

When attempting an upgrade from NSX 4.1.1 to 4.2.3.2, the Upgrade Coordinator may stall or fail at the Host Upgrade stage. This specifically affects hosts that were previously part of a vCenter cluster but have since been removed.

The NSX Manager UI displays a Cluster Name in the node name fields for a host that is currently removed from VC under Standalone Transport Node.

The Transport Node NSX configuration remains in a  "Orphaned" state.

Communication between the Manager and the Transport Node fails with the following RPC error:

Host configuration: Failed to send the HostConfig message. [TN=TransportNode/cc####-####-####-####-########]. Reason: Failed to send HostConfig RPC to MPA node:cc####4-####-####-####-########. Error: Unable to reach client ########-####-####-####-########, application SwitchingVertical. LogicalSwitch full-sync: LogicalSwitch full-sync realization query skipped.

Environment

VMware NSX

Resolution

To resolve this synchronization mismatch, the cache of the Management Plane must be refreshed. In this case, restarting the NSX Manager node's resolved the issues