Unable to storage vmotion VMs from a VMFS datastore

Products

• VMware vCenter Server

Issue/Introduction

Symptom:

• VMs on the datastore are showing up as inaccessible and orphaned state. 

• Storage vmotion fails with an error "Cannot complete file creation operation."
  
  

  

Validation:

• VMs seen from the vCenter on the datastore are just stale entries. There are no VM files on the datastore.
  
  SSH to the ESXi host and navigate to the VM directory. No files are present on the datastore -
  
  [root@esxihost01:/vmfs/volumes/<UUID of Datastore>/<VM Folder>] ls -l
total 0

Environment

VMware vSphere ESXi 7.x

VMware vSphere ESXi 8.x

Cause

• VM is unable to access the storage since the host has lost access to the LUN.

   

• This issue occurs when partition tables for VMFS volumes are removed while ESX/ESXi is using them. This can result from other systems clearing partition tables from visible LUNs to prepare them for use.

For example:

• • During installation of a new operating system on a physical server, it may delete or overwrite partitions on those LUNs as part of the installation.
    
    
  • After an operating system is installed, it can be directed to overwrite partitions of LUNs it has visibility to.
    
    
  • The ESX installer can be directed to overwrite an existing VMFS Datastore, though it will warn against possible data loss.
    
    
  • Unattended or automated installations, such as kickstart, jumpstart, and FAI, or full-disk imaging processes may overwrite contents of LUNs that they have access to.
    
    
  • Utilities such as fdisk, parted or dd may be invoked on any server with LUN visibility, and remove or alter the partition table on the LUN.
    
    
  • Storage array issues may impact contents of LUNs.

These or similar situations may accidentally occur if LUNs are presented to a mixture of ESX/ESXi and non-ESX/ESXi servers. To prevent against the above situations, restrict LUN visibility to only servers which are trusted and which require access to the VMFS Datastores.

• In "/var/run/log/vmkernel.log", we see below entries -
  
  2025-05-03T18:17:24.759Z cpu42:2100095)WARNING: Partition: 1403: Partition 1 is active, failure to update partition table for eui.################################


• In "/var/run/log/vobd.log", we see below entries -
  
  2025-05-03T19:45:17.344Z: [scsiCorrelator] 5679137454718us: [esx.problem.scsi.device.state.permanentloss] Device: eui.################################ has been removed or is permanently inaccessible. Affected datastores (if any): Unknown.
  

Resolution

• Recover VMs from Backup.  

Additional Information

Partition table fails to update active partitions