I'm working on a disaster recovery plan for my company and am trying to script out a recovery scenario for the event that we have to restore our system to a remote location and all that we have are our backup tapes and an empty server. I'm testing the recovery plan onsite isolated from my production network. My "empty server" is running VMWare ESXi 4.1 and I am currently trying to restore my primary NetWare file server.

I already have a Windows server built and running the Backup Exec Console and have the NetWare server built nearly identical to the production server, the only exceptions are processors and storage capacity, but otherwise it has the same name, same IP address and same services installed. So once I have my skeleton of a server built and the BE Remote Agent talking to the console, I'm trying to do a restoration. I have no trouble restoring my file system back to the server, where I am running into trouble is with the Novell Directory Root and Schema objects. In short, I keep getting an "Access is Denied" error from the restoration process, but it isn't quite that simple. I am using the Admin account to do the restoration, so it should have all the rights needed to restore. Next, I keep experiencing an abend related to the certificate services when the restoration runs, and thirdly, large parts of my tree do get restored but some OU's will have nearly every user flagged as an unknown object with the round yellow question mark icon, where other OU's restore perfectly.

On my restoration I am telling Backup Exec to overwrite all existing files as opposed to the "unless newer" option because I'm currently testing the restoration off of last Friday's backup tapes, and I only built the NetWare recovery server yesterday, so the server's files will end up being newer than what is on tape anyway.

Does anyone know how to properly restore the Root and Schema files using Backup Exec (I am running the latest version 2010) for a bare metal tape restoration? I did see recommendations to create a new server in my Recovery environment that was connected to the production environment so that it obtains a proper replication and I won't be having this trouble anyway. The only problem with that method is that it assumes that for recovery in a true disaster that the replicated server is available to reconnect to, which might not be the case in a real scenario. Thanks again for any advise!