WSFC Cluster VM Fails to Power On with "Unable to access file... locked file/disk" Error when using Physical RDMs
search cancel

WSFC Cluster VM Fails to Power On with "Unable to access file... locked file/disk" Error when using Physical RDMs

book

Article ID: 426859

calendar_today

Updated On:

Products

VMware vSphere ESXi VMware vSphere ESX 8.x

Issue/Introduction

A Virtual Machine (VM) that is part of a Windows Server Failover Clustering (WSFC) configuration and uses physical Raw Device Mappings (pRDM) fails to power on.

The VM presents an error message indicating that a file is locked or cannot be accessed, similar to the following:

Unable to access file /vmfs/volumes/#########-########-####-###########/<VM-Folder>/<VM-name>_6-rdmp.vmdk since it is locked

Environment

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

Cause

The shared physical mode Raw Device Mappings (pRDM) are inadvertently configured on the Virtual Machine's SCSI Controller 0.

This specific controller does not have physical bus sharing enabled, which is required configuration for sharing pRDMs across multiple nodes in a Windows Server Failover Cluster (WSFC).

This misconfiguration leads to a file lock conflict, preventing the VM from powering on.

Resolution

To resolve the issue, reconfigure the affected WSFC node's VM settings to move the shared pRDMs to a SCSI controller that has been correctly configured for sharing.

Steps:

  1. Edit the VM Settings of the affected WSFC node.

  2. Identify the shared physical mode RDMs that are currently configured on SCSI Controller 0. (Non shared disks)

  3. Ensure Consistent Configuration: Verify that the virtual disk SCSI IDs are consistent across all VMs hosting nodes of the same WSFC.

  4. Verify and Configure Sharing:
    • Check for another SCSI Controller (e.g., SCSI Controller 1).
    • Confirm that its SCSI Bus Sharing setting is set to Physical.
    • If a correctly configured controller does not exist, add a new SCSI controller and set its sharing mode to Physical.

  5. Re-attach pRDMs: Edit the settings for the shared pRDMs and attach them to the correctly configured SCSI Controller (e.g., SCSI Controller 1) where SCSI Bus Sharing is set to Physical.

  6. Power on the WSFC node VM. The VM should now power on without the "locked file/disk" error.

Additional Information

Microsoft Windows Server Failover Clustering (WSFC) with shared disks on VMware vSphere 8.x: Guidelines for supported configurations

Powered by Wolken Software