This topic describes common workarounds of issues that you might encounter while backing up or restoring Hyper-V virtual machines.
Linux virtual machine boots into maintenance mode after restore
If the disk sequence changes after restore, the boot order changes for the Linux virtual machine and the virtual machine boots into maintenance mode or some virtual disks are not visible.
For the Linux virtual machine, change the disk sequence in /etc/fstab. The virtual machine boots without any problem after you enter the correct disk sequence. For more information, see RedHat documentation. [External link]
Virtual machine restored to an alternate host fails to start
When VM is restored on another host, sometimes it fails to start with an error: 'Unnamed VM' could not initialize. (Virtual machine ID).
The event log shows the error: The security ID structure is invalid (0x80070539).
On the Hyper-V host:
- Open an administrative PowerShell command window.
Grant-VMConnectAccess -VMName "Name of VM that is not starting" -UserName "Domain and username of the current user"
For more information, see Virtual machine cannot start - The security ID structure is invalid (0x80070539) [External link to a Microsoft support article.]
Cannot perform a file-level restore from a virtual machine backup
One of the reasons why a virtual machine FLR fails is the presence of user snapshots. If user snapshots are present, Druva Phoenix backs up a virtual machine, however, FLR from such virtual machine backups is currently not supported. The backup job for such a virtual machine is marked as successful with errors.
Delete the user snapshots and backup the virtual machine to use file-level restore.
The following error messages are encountered at the time of virtual machine restore:
- Failed to add VHDs to VM dev03 (#10009000b)
- General access denied
This error can occur if you are restoring a virtual machine or a virtual disk to a Windows Server 2008 or Windows Server 2008 R2 host with SMB share target, and appropriate write issues on the share are not available.
To resolve this issue:
- Flush and purge the existing authentication tickets between the host and the share, and
- Re-apply write permissions to the target folder
- Hyper-V host is Domain\hypervhost1
- SMB Share system is Domain\smbsharesystem
To flush and purge the existing authentication tickets:
Run the following commands as administrator on the Hyper-V host:
klist tickets klist purge
Create a new folder and apply write permissions to the target folder
Run the following commands as administrator on the Hyper-V host
MD C:\SomeFolder ICACLS C:\SomeFolder /GRANT Domain\Administrator:(OI)(CI)F /GRANT Domain\hypervhost1$:(OI)(CI)F Domain\smbsharesystem$:(OI)(CI)F NET SHARE SomeFolder=C:\SomeFolder /GRANT:Domain\Administrator,FULL /GRANT:Domain\hypervhost1$,FULL /GRANT:Domain\smbsharesystem$,FULL
Note: Use the above example as a guide. Replace the inputs SomeFolder, hypervhost1, and smbsharesystem based on your setup.
Restore to an alternate path is successful but additional disk drives are missing, failed, or offline.
This issue can occur when you are restoring virtual machines to Windows 2008 server, and you have multiple virtual disks of different types attached to it, such as simple, spanned, striped, or mirror disk.
In such a case, after you restore the virtual machine and turn it on, you can only see the primary disks in the virtual machine.
Launch the DISKPART utility or the disk management console (diskmgmt.msc), select each disk individually, and bring each virtual disk online manually. For example, run the following commands in the DISKPART utility:
DISKPART> select disk 1 Disk 1 is now the selected disk. DISKPART> DISKPART> ONLINE DISK DiskPart successfully onlined the selected disk.
Backup fails with the HYPERV10 error after re-registering a Hyper-V host
This issue can occur when you generate a re-registration activation token for one host and use the token to re-register another host. For example, you have two hosts registered with Druva Phoenix for backup, hosts h1 and h2. Host h1 has virtual machines a1, a2, and a3 running on it, and host h2 has virtual machines b1 and b2 running on it.
If you generate a re-registration token for the host h2, but install and activate the agent using the re-registration token on host h1:
- Backup jobs for virtual machines b1 and b2 fail since the host h2 is still disconnected
- Backup jobs for virtual machines a1, a2, and a3 fail since the host h1 is activated, however, the agent is looking for virtual machines b1 and b2 on the host
- Host h1 is listed as host h2 on the Hyper-V page of the Phoenix Management Console in the organization it was registered
- There are two hosts listed as host h2 on the Hyper-V page of the Phoenix Management Console in the organization the two hosts were registered
In this example, there are two hosts that are listed as h2, however, the host that is still connected to the Druva Cloud is host h1. To resolve this issue, re-register the host h1 that is listed as host h2 with a new re-registration token. After you re-register the host, you can see that the host h1 is now listed correctly on the Phoenix Management Console. After you re-register the host h1, select the disconnected host h2 and re-register it with a new re-registration token. After you re-register the host h2 with the correct activation token:
- Both the hosts are connected to the Druva Cloud
- Both the hosts are listed properly with the backup jobs running as expected
Failed to create application-consistent backup snapshots on Hyper-V host Windows Server 2019 with the guest operating server Windows Server 2012
The creation of the production checkpoint on Hyper-V Guest OS Windows Server 2012 fails for the older version of Integration Services. For older version of Integration Services, the Progress logs shows the following details:
Failed to take app consistent checkpoint. Trying crash consistent checkpoint
To ensure that your backups are application-consistent, download and install updates on the Windows update page of the Hyper-V host.