Skip to main content

Recently upgraded a client to a new server, running ESXi. It has 2 windows servers and a win10 backup and replication server as VMs. The old server was also ESXi only running 1 windows server and the windows agent. When I first tried to bring the new server up starting with a bare metal restore of the original server I ran into problems because Veeam would not restore the server to a functional state because it was a standalone domain controller. There were many suggestions on how to correct this but none of them worked. Finally I got around this by setting up a new windows server on the new hardware with the old server still running on the old hardware, and made it a BDC. Then when I turned off the old server and restored it to the new hardware it synced to the existing Domain Controller and worked normally.

Now I want to take backups from the same vm on the new server, and restore to the original vm on the old hardware, which will be in a different location, creating a disaster recovery machine of their accounting system in case they need it. The problem comes back to restoring a standalone server that is also a domain controller.

Previously I was doing a full bare metal recovery of the server … Is there some other type of restore I could use to restore the system drive, perhaps a volume restore of only drive c: ?

Or maybe step by step instructions on restoring a stand alone domain controller where there isn’t a a second domain controller to verify the domain data is up to date? or maybe a hidden option to tell veeam that “yes I know it’s a domain controller … go ahead and use it”?

Thanks for your help, Cla.

Hi ​@clabrown , in a single DC scenario, or if you want to test the restore in an isolated environment you have to perform an authoritative restore of you domain controller.

Read this useful KB https://www.veeam.com/kb2119

If you need more information please ask without any problem.


Hi,

Two scenarios come to my mind:

  1. Using Veeam replication to have a ready-to-use DC VM on a second target
  2. Creating a second VM and configure it as second DC so that the environment has 2 DCs in case that is possible

I also agree with ​@marco_s - you should use AAP and perform an authoritative approach in any case.

 

Keep in mind the following: When you “manually” do things within the restore the DC might run into problems because there are timestamps, DNS services and IP mappings with dependencies to the services.

If you use an agent restore you have to think about drivers and performing “afterworks” after restoring depending on the configuration.


Comment