Power on failure for VM with RDM disk on WSFC node

Products

• VMware vSphere ESXi

Issue/Introduction

• When powering on a virtual machine configured with a RDM disk, task fails with the following error message:
  
  "File system specific implementation of Ioctl[file] failed Failed to start the virtual machine. Module 'Disk' power on failed. Cannot open the disk '/vmfs/volumes/vm file path/vmname.vmdk' or one of the snapshot disks it depends".


• Logs indicate RDMs are incorrectly identified as VMFS datastores or report reopening failures.
  
  /vmfs/volumes/<datastore uuid>/<VM Name>/vmware.log:
  
  YYYY-MM-DD:T:HH:MM:SS In(05) vmx - [msg.migrate.resume.fail] The VM failed to resume on the destination during early power on.
YYYY-MM-DD:T:HH:MM:SS In(05) vmx - [msg.moduletable.powerOnFailed] Module 'Disk' power on failed.
YYYY-MM-DD:T:HH:MM:SS In(05) vmx - [msg.disk.noBackEnd] Cannot open the disk '/vmfs/volumes/<datastore uuid>/<VM Name>/<RDM Disk>.vmdk' or one of the snapshot disks it depends on.
YYYY-MM-DD:T:HH:MM:SS In(05) vmx - [msg.fileio.lock] Failed to lock the file
YYYY-MM-DD:T:HH:MM:SS In(05) vmx - [vob.fssvec.OpenFile.file.failed] File system specific implementation of Ioctl[file] failed
  

  /var/run/log/hostd.log

  YYYY-MM-DD:T:HH:MM:SS Er(163) Hostd[2#####6]: [Originator@6876 sub=Vdisksvc opID=<OPID>  sid=52####81 user=vpxuser:DOMAIN\Administrator] ValidateRawDiskDevice: Device vml.##### used for vmfs datastore
YYYY-MM-DD:T:HH:MM:SS In(166) Hostd[2#####6]: [Originator@6876 sub=Vmsvc.vm:/vmfs/volumes/<DATASTORE UUID>/<VM NAME>/VMNAME.vmx opID=<op id> sid=5###1 user=vpxuser:DOMAIN\Administrator] ValidateRawDiskDevice: Rdm device is not supported

Environment

VMware vSphere 8.x

Cause

The issue occurs when the RDM pointer file or the underlying device is incorrectly flagged as a VMFS datastore.

Resolution

To resolve this issue, remove and recreate the existing RDM mapping (pointer) files. This process destroys the descriptor file but does not affect data on the actual LUN.

1. Identify the problematic RDM disk displaying the warning 'Device vml.##### used for vmfs datastore' as referenced in Identifying Virtual Machines with Raw Device Mappings
   
   
2. Remove Existing RDM Pointer Files:
   
   
   1. Right click the affected VM in the vSphere Client inventory and select Edit Settings.
   2. Expand the affected RDM hard disk and document the Target LUN, VML ID, Compatibility Mode, and SCSI Controller settings.
   3. Click the X icon next to the RDM hard disk to remove it from the Virtual Machine.
   4. Select Delete from disk (Note: This destroys the mapping file, but not the LUN content.).
   5. Click OK. 
      
      
3. Recreate RDM Pointer Files:
   
   
   1. In Edit Settings, click Add New Device and select Raw Device Mapping
   2. Choose the appropriate LUN from the list based on the ID noted earlier.
   3. Select a datastore for the LUN mapping file (usually the VM's home directory).
   4. Select the appropriate Compatibility Mode (Physical or Virtual).
   5. Ensure the SCSI Controller and Sharing settings match the original configuration.
   6. Click OK and attempt to power on the VM.

If the issue persist, proceed to engage Broadcom Support to investigate the issue further.

Additional Information

Reference:
Moving an existing RDM to a different virtual machine.
Add an RDM Disk to a Virtual Machine
Add new RDM to existing Microsoft Windows Server Failover Clustering (WSFC) VMs