Skip to main content

How to migrate ESX VM with independent disks and PCIE to AHV

Nutanix Move doesn’t support VM with Independent disk and PCIe pass-through device. So you can use manual method to migrate the VM if has Independent disk or PCIE pass-through device. Please refer following article for manual migration: Transferring Virtual Disks to an AHV Cluster


For Independent Disk:Independent disks are not supported with Move until you remove those disks. To  fully migrate  a VM that has these disks? Remove the independent disk, migrate the VM with move, then manually migrate the data from the independent disk and add the disk back to the migrated VM. The challenge is that CBT APIs do not work with independent disks.  You can alternatively migrate the data to regular VMDKs first and then attempt the migration with Move.


From Move 3.6 guide

Unsupported features:This topic lists the unsupported features for migration from ESXi to Nutanix Cluster on AWS and AHV.

  • Guest operating systems not supported by AHV.
  • PCIE pass-through
  • Independent disks
  • VMs with multi-writer disks attached
  • VMs with 2 GB sparse disk attached
  • VMs with SCSI controllers with a SCSI bus sharing attached

    Note: Change SCSI bus controller to None.

  • Migration of Windows VMs with dynamic disks
  • Migration of VMs from ESXi standalone hosts with free license

Yes Exactly , If you read my previous comments , i put a statement “Independent disks are not supported with Move until you remove those disks” If you have a vm with Independent disk then pls follow below steps “You can alternatively migrate the Independent disk data to regular VMDKs (Dependent Disk) first by manual approach (Move doesn't support this feature)and then attempt the migration with Move.


You can manually turn them into normal disks (I think, VM power off may be required) and then perform the data seeding and then migration. You can turn the disk into Independent again after the migration. Data seeding is based on a snapshot. Independent disks are excluded from the snapshot but are present (referred to) in the VM configuration file. Hence when attempting to re-create the VM the task is destined to fail. That’s why it is unsupported.