ESXi host restarts unexpectedly during TSDM remediation

Products

• VMware vCenter Server
• VMware vSphere ESXi

Issue/Introduction

• ESXi host restarts unexpectedly during during a TSDM (Dell Transparent Snapshot Data Mover) remediation whose only change is a live-installable software component.
• A cluster managed by a vLCM cluster image that includes a Firmware and Drivers Addon from HPE OneView for VMware vCenter HSM (Hardware Support Manager).
• The HSM task info JSON string in /var/log/vmware/vmware-updatemgr/vum-server/vmware-vum-server.log shows the below: 

  "fwCompliance": "COMPLIANT",
  "fwStageStatus": "NOT-STAGED",
  "impact": {
      "finalRebootHwResetRequired": false,
      "finalRebootRequired": false,
      "intermediateRebootHwResetRequired": false,
      "intermediateRebootRequired": true,
      "maintenanceModeRequired": false,
      "postImageUpdateRequired": true,
      "preImageUpdateRequired": true
  }

• TSDM upgrade attributes shows:

  Maintenance Mode Required: False
  Live Install Allowed: True

Environment

VMware vCenter Server

VMware vSphere ESXi

Cause

In the HSM Task Info JSON, the "finalRebootRequired" is set to "false" which means no final reboot is required after the firmware is installed, however, the "intermediateRebootRequired" field is set to "true" which specifically indicates that a reboot must occur in the middle or prior to the update sequence, either because it is required as a step in the middle of the firmware upgrade.

The reboot is being triggered and requested by the HPE OneView Hardware Support Manager (HSM) rather than the TSDM component itself. In a vLCM image-managed cluster, vLCM queries the HSM for the "impact" of any change to the cluster image, and the snippet you provided confirms that the HPE HSM is reporting "intermediateRebootRequired": true during the remediation sequence.

Resolution

The ESXi reboot is expected as per the HSM Task Info JSON.