ESXi Hosts Showing “Host Disconnected” in NSX within the VxRail cluster
search cancel

ESXi Hosts Showing “Host Disconnected” in NSX within the VxRail cluster

book

Article ID: 422429

calendar_today

Updated On:

Products

VMware NSX

Issue/Introduction

  • Following a recent manual certificate replacement (performed because the CARR script was not functioning at the time), all ESXi hosts in a VxRail cluster began showing a “Host Disconnected” status in the NSX UI.

  • The Host Disconnected error details for each Transport Node reported:

    • “Heartbeating between host and manager is down.”

  • The issue affected all hosts in the cluster simultaneously, rather than a single Transport Node.

  • Attempts to resolve the errors from the NSX UI failed immediately with no change in host status.

  • Network connectivity between ESXi hosts and NSX Managers was verified and confirmed healthy:

    • ICMP ping successful in both directions (hosts ↔ managers)

    • Netcat connectivity tests from the Host Transport Node CLI were successful on required ports:

      • nc -zv <NSX_Manager_IP/FQDN> 1234

      • nc -zv <NSX_Manager_IP/FQDN> 1235

  • These checks confirmed that the issue was not related to network connectivity, but instead pointed to a trust or certificate-related problem between NSX and the hosts.

Environment

VMware NSX 4.1.0.2

Cause

The issue was caused by a thumbprint mismatch between the registered Compute Manager (vCenter) and NSX, along with stale certificate entries remaining from the prior manual certificate replacement.

Resolution

Workaround:

  • Download the latest CARR script using the Using Certificate Analyzer, Results and Recovery (CARR) Script to fix certificate related issues in NSX documentation.

  • Secure copy (SCP) the CARR script to the root directory of any NSX Manager, then extract and execute the script following the steps outlined in the same documentation.

  • Review the CARR script results and remediate all identified certificate and trust-related issues, including any Compute Manager thumbprint mismatches or stale certificates.

  • Once all fixes have been successfully applied, navigate to the NSX UI, select each host showing a Host Disconnected status, click View error details, and resolve the reported errors.

  • Confirm that all hosts transition to SUCCESS and UP status in the NSX UI.

 

Additional Information

 

  • Manual certificate replacement can leave stale certificate data and thumbprint mismatches if NSX is not updated accordingly.

  • The CARR script is the recommended tool for detecting and correcting certificate, trust, and thumbprint inconsistencies between NSX and vCenter.

  • When encountering widespread Host Disconnected states with heartbeat errors, and network connectivity is confirmed, certificate trust and Compute Manager thumbprints should be validated early in the troubleshooting process.