Default alarm "vSphere HA virtual machine monitoring error" is triggered during guest OS-initiated shutdown

Products

• VMware vSphere ESXi
• VMware vCenter Server

Issue/Introduction

When performing a normal shutdown of a virtual machine from within the guest OS (such as using OS shutdown operations or commands), a temporary processing race condition may occur between vCenter Server and ESXi, triggering the following default alarms and logging errors:

• Alarm Display in vSphere Client
  • Triggered Alarm: vSphere HA virtual machine monitoring error
  • Description: vSphere HA cannot reset [VM Name] on [ESXi Host Name] in cluster [Cluster Name]
• Error Outputs in Various Log Files

hostd.log:

YYYY-MM-DDT01:17:58.605Z Wa(164) Hostd[...]: [...] Query VMX about hlstate failed Fault cause: vim.fault.InvalidPowerState
YYYY-MM-DDT01:21:01.928Z Er(163) Hostd[...]: [...] Invalid transition requested (VM_STATE_OFF -> VM_STATE_RESETTING): Invalid state

fdm.log:

YYYY-MM-DDT01:21:01.936Z Wa(164) Fdm[...]: [...] Error while resetting VM /vmfs/volumes/.../VM_Name.vmx: N3Vim5Fault17InvalidPowerState9ExceptionE(Fault cause: vim.fault.InvalidPowerState
YYYY-MM-DDT01:21:01.936Z In(166) Fdm[...]: [...] Reset no 0 failed for VM /vmfs/volumes/.../VM_Name.vmx
YYYY-MM-DDT01:21:01.936Z Db(167) Fdm[...]: [...] New event: Event=vim.event.VmDasResetFailedEvent vm=/vmfs/volumes/.../VM_Name.vmx host=host-...

Note: This issue does not occur when the shutdown is initiated from the "Shut Down Guest OS" menu in the vSphere Client.

Note: The virtual machine itself eventually transitions to a healthy `Powered Off` state, and no data corruption or functional impact on the system occurs.

Environment

VMware vSphere ESXi 7.0x

VMware vSphere ESXi 8.0x

VMware vCenter Server 7.0x

VMware vCenter Server 8.0x

Cause

This issue is an expected behavior (race condition) caused by a timing conflict between the vSphere HA "VM Monitoring" automated recovery process and the successful completion of the guest OS shutdown.

Resolution

This issue represents a temporary timing conflict (race condition) that occurs when the vSphere HA protection logic operates exactly as designed. It does not indicate a system malfunction or defect.

The virtual machine has successfully completed its shutdown, and there is no impact on data or configuration file integrity. Therefore, this alarm and error event can be safely ignored.