Hyper-V pass-through disks are not accessible after system is restarted

Applies to: Hyper-V Server 2008Microsoft Hyper-V Server 2008 R2

Symptoms


Consider the following scenario:

You have a Hyper-V Host, which is connected to SAN storage, where the disks which are connected via Fiber Channel or iSCSI and are configured as Hyper-V pass through LUNs to make them accessible directly to the Hyper-V guests.

In this scenario, when the host machine is rebooted, some or all of the pass-through disks may be inaccessible, and the guests are unable to boot if the LUN was a boot LUN, or unable to access data where it was a data volume.

Cause


This behavior can occur when all of the following happen:
  • Hyper-V is configured to pass through LUNs to Hyper-V guests.
  • A change on the array changes the PNP instance ID of the device, making it appear as a new device.
  • As a result of the PNP instance ID changing, the disk order changes.
  • As a result of the disk order changes, the LUNs can no longer be passed through.
Conclusively determining the cause of this can be difficult. In order to determine if the disk ordering changes were the result of a PNP instance ID change it would be necessary to have access to registry data from a backup of the machine in question before the issue occurred to compare the data with the current copy.

For example, take the case below

Note:  the GUID and vendor ID have been modified to show a generic example, where on two different dates, the same GUID has surfaced as different device ID’s.

Current version (below)
 

SCSI\DISK&VEN_DISK_DEVICE&\{11111111-1111-1111-1111111111111111}&DISK_DEVICE43

SCSI\DISK&VEN_DISK_DEVICE&\{11111111-1111-1111-1111111111111111}&DISK_DEVICE45

SCSI\DISK&VEN_DISK_DEVICE&\{11111111-1111-1111-1111111111111111}&DISK_DEVICE47

 
Original version from backup (below):
 

SCSI\DISK&VEN_DISK_DEVICE&\{11111111-1111-1111-1111111111111111}&DISK_DEVICE17

SCSI\DISK&VEN_DISK_DEVICE&\{11111111-1111-1111-1111111111111111}&DISK_DEVICE15

SCSI\DISK&VEN_DISK_DEVICE&\{11111111-1111-1111-1111111111111111}&DISK_DEVICE13

 

Resolution


To resolve this issue contact your storage hardware vendor, to determine the cause of PNP instance ID changing. The most common occurrence of this inadvertent change is during a firmware update to an array or array controller.

More Information


For Microsoft Windows Server 2008 R2, the PNP Instance ID was introduced as a method for maintaining disk ordering across reboots. This functionality depends on the device continuing to appear as the same device across reboots. If the PNP Instance ID changes as a result of a hardware change, then it makes the LUN surfaced to Windows appear as a new device, and the disk order is not maintained.