NBU SharePoint GRT Restore - Database System Error

[SYMAJ]

I have trawled through the many technotes on this subject, which seems to be a common cause for confusion, and still cannot progress past this error.

I have a simple Sharepoint 2007 setup as follows:  1 x WEB server, 1 x APPS server, 1 x SQL server.  All servers W2K8R2, SQL 2K8R2.

I have done the following in preparation for the GRT Restore:

1. Install NBU 7506 client on each of the three SP servers

2. Set ALL NBU services on the 3 SP servers to run under an account which is Domain Admin, Local Admin, Sharepoint Admin and SQL sysadmin.

3. Set the SP credentials within the Windows tab of the client properties for all 3 SP servers to run as the SharePoint admin.

4. Installed NFS on the SQL server and disabled the NFS Server Service. (NFS is installed on the Master Server, Media Server is a 5230 - Linux)

5. Set up a VMWARE policy, specified all 3 SP servers (by query) and checked the ENABLE SHAREPOINT RECOVERY box in the VMWARE tab, Enable BLIB is un-checked.

6. Distributed Application Restore Mapping has been set with two entries:  WEB to APP and WEB to SQL

Backups run fine, and I can restore a full VM image or a file level restore no problem.

I am having difficulty when I try and restore Sharepoint Resources (MS-Sharepoint policy type).  When I select the APP server as source and target (or the app server as source and web as target) I can drill down through the SharePoint resources through the resources and folders.  When I get to the 'tape drive' icon and try and expand it then I get the DATABASE SYSTEM ERROR.  A restore job runs successfully in the Activity Monitor, even though I get the error in the GUI.  There is no specific error message - just the DATABASE SYSTEM ERROR.  Which logs will be of help here ?

I appear to be going around in corcles at this point, and would appreciate any advise as to what I may have missed......

Thanks,

AJ

[SYMAJ]

Since the original post I have made the following changes based upon the documentation:

NBU Client Service and NBU Legacy Client Service now running as LOCAL SERVICE on each of the 3 Sharepoint Servers - NO CHANGE

Change Distributed Application Restore Mapping to be APP - SQL and APP - WEB (based upon it being the APP server I have to specify in the BAR Restore Client) - NO CHANGE

As I am still getting the DATABASE SYSTEM ERROR I would appreciate any advice as to what I am missing here.

Thanks,

AJ.

[ChipApple]

AJ,

I just received the same error message "ERROR: database system error" except I am receiving the error when I am trying to view an Active Directory GRT backup to do a restore. I am not even getting to the restore option. The backup ran fine with a STATUS 0. I am assuming the error message has something to do with the NBU database and possibly relating to nbfsd would be my guess.

I am opening up a Support case with Symantec. We just upgraded to NBU 7.5.0.6. I will keep you posted.

Best Regards,
Chip

[Umair_Hussain]

Chip, Correct me if i am wrong, i think its a known error and etrack ET2708814 is associated to database system error when browsing restore for GRT backups.

[ChipApple]

Hey BIAB,

Thanks for the Etrack number listed above but in my case, it does not apply. This is a AD GRT backup/restore on a physical system, but the issue might be the same.

[SYMAJ]

My issue turned out to be related to NFS and my lack of understanding of the SharePoint environment.  I assumed (incorrectly) that when it referred to the 'front end server' it meant the WEB server.  I had NFS installed on the SQL server, WEB server, NBU Master Server and Media Server - the only server which didn't have it was the APPS server.

I raised a call with Support but whilst waiting for them to call back spotted this and thought it worth a try to install NFS on the APPS server.  I did and all was resolved.  It takes approx 1 minute to expand the backup image when in the BAR GUI but all works OK.

AJ.

[ChipApple]

Hey AJ,

Can you actually perform a restore now that you can see the Sharepoint objects? Just curious. We have also been able to produce the issue with Exchange 2010 GRT as well but the issue seems more intermittent and random. Strange...

~Chip

[ChipApple]

Ok we are getting closer to a solution. Turns out we are  using SLP going to a DataDomain as our Storage target for GRT backups. This was affecting NBU GRT restores (AD, Exchange) to see the granular objects. and causing the database system error. Note: all GRT backup jobs were completing with a STATUS 0 under SLP. What I have done since then is point the Storage target straight to the Data Domain instead of the SLP and this gives me the ability to browse the GRT objects.

I have opened up a case with Symantec and we will be reviewing the SLP configuration.

Stay tuned for more details.

~Chip