HCX Interconnect (IX) tunnel status persists as "Unknown" following appliance re-deployment

Products

• VMware HCX

Issue/Introduction

Following a redeployment of an HCX Interconnect (IX) appliance, the Service Mesh UI displays an "Unknown" status for the appliance. Specific indicators include:

• Overall transport tunnel status: Unknown
• Overall encryption tunnel status: Unknown
• Service pipeline status: Unknown
  
  
  
• However, the IX tunnel status from the target site showed as "Up".
• app.log at /common/logs/admin
Error: Failed to configure Mobility Agent host. Reason: No route to host (Host unreachable)
• In affected environments, the ccli output reveals two IX appliances registered to the HCX Manager for the same site pairing:
  
  
• Although the transport tunnel may show as Up within ccli (via show transport status), the HCX Manager UI cannot reconcile the duplicate registration, leading to the "Unknown" state.
  
  

Environment

VMware HCX

Cause

This issue is caused by a "stale registration" in the HCX Manager database. Only one IX appliance can be registered per Service Mesh. If a previous Service Mesh was not deleted cleanly (removed from the UI but failed to unregister the appliance), the database retains the old record.

When the new IX is redeployed using the "same IP address" as the stale record, the HCX Manager experiences a conflict between the Disconnected stale node and the Connected active node, preventing proper status reporting in the vCenter plugin or HCX UI.

Resolution

To resolve this conflict and restore service visibility, implement one of the following options:

Option 1: Redeploy with a Unique IP (Recommended)

Avoid IP reuse to bypass the database conflict:
1. Edit the Network Profile used by the IX appliance to provide a new, unused IP address.
2. Redeploy the IX appliance within the Service Mesh using the Force Update or Redeploy option.
3. Verify that the new appliance registers with a unique IP that does not match any stale entries in ccli.

Option 2: Database Cleanup

If the IP address must be reused, a manual cleanup of the stale registration is required:
1. Open a "Support Request" with Broadcom Technical Support - Creating and managing Broadcom cases
2. Reference this article and provide the output of the ccli list command.
3. Support will assist in manually pruning the stale Disconnected entries from the HCX Manager database to allow the active appliance to report status correctly.

IMPORTANT: Do not attempt to manually modify the HCX Manager database without guidance from Support, as this can lead to Service Mesh corruption.