Monitoring tool reports PDL with storage UUID

Products

• VMware vSphere ESXi

Issue/Introduction

Symptoms : 

•  The Monitoring tool reports a PDL (Permanent Device Loss) error for the volume with the "volume UUID."

Validation Step: 

• Log in to the monitoring tool to investigate the cause of the PDL error.
  
  
  
  Based on the screenshot above, it has been confirmed that the setting "VMkernel.Boot.terminateVMOnPDL" should be enabled to prevent PDL errors.

Environment

VMware vSphere ESXi 6.x
VMware vSphere ESXi 7.x
VMware vSphere ESXi 8.x

Cause

• The script is designed to report a PDL error if the setting "VMkernel.Boot.terminateVMOnPDL" is configured as false.
• Since this setting was set to false, the tool reported the PDL error.

Cause validation:

• Login into vcenter. Navigate to host > configure > Advanced System Settings. Check the value of the "VMkernel.Boot.terminateVMOnPDL" key.
  
  
  As shown in the screenshot above, the key "VMkernel.Boot.terminateVMOnPDL" is set to false, which triggered the PDL error in the monitoring tool.

Explanation of the Setting:
Datastore with PDL failure response: When a PDL event occurs, vSphere HA will terminate all affected VMs and attempt to restart them on other hosts that still have connectivity to the datastore. This setting is configured at the cluster level, ensuring that all hosts are monitored and managed by HA.
VMkernel.Boot.terminateVMOnPDL: This ESXi setting assists vSphere HA in performing the action described above. By default, this setting is set to false, meaning that if a PDL occurs, the VMs will not be terminated and restarted by HA. Setting this to true allows the ESXi host to terminate VMs during a PDL event, enabling HA

 

Resolution

To ensure that vSphere HA terminates and restarts VMs on other hosts during a PDL event, set the "VMkernel.Boot.terminateVMOnPDL" parameter to true on all ESXi hosts in the cluster.