SQL 2012 Always On Recovery Failure

Question

Hello, I've been successful in getting our SQL 2012 Always On Availability group to be backed up via our Networker Server (8.2.1.0 681) to a standard AFTD Device, using NMM 8.2.1.

When I try to do a recover following the document "docu57655_NetWorker-Module-for-Microsoft-for-SQL-VDI-Release-8.2-SP1-User-Guide.pdf" regarding Federated Recovery, I do not get an option to recover a single Database (I receive "This item is part of SQL 2012 Always On group and can't be marked) so i have to recover the entire set of Databases in the Availability group, which isn't what i want to achieve.

When i try a different method of restoring, via Restore\Copy and selecting my Availability Group as the source, I can then select an individual database for recovery, and am permitted to select the option to Recovery to a different database, and different file paths etc, but when i undertake the restore, it starts ok, and i see a Browse session on the Networker Console, but I receive the errors;

NSR critical 34 Error opening tape session for %s. 1 0 16 MSSQL:/test_recv

NSR warning 34 VDI device did not close properly. 0

xbsa.messages shows;

_nwbsa_open_restore_session: System detected error due to: Failed to establish restore session. Exiting. Operation aborted

So, couple of questions are;

1. Can i do a selective restore of an individual database when backing up an Availability Group.

2. Why am i seeing the VDI Device error when trying to recover via Restore\Copy method?

Kind Regards

Adrian

Message was edited by: AdrianSahota

ble1 · Answer

I assume you use domain user used for backup when doing restore or is it DBA user?  If DBA user, is it listed in NW roles? I'm not 100% sure, but since AoAG backup is normally done via cluster name, I suspect you always need to use copy approach as otherwise restore window would just show system dbs locally (versus user dbs).  With that said, verify that remote access is also granted correctly.

ble1 · Answer

Must be something silly... since most likely ddboost is used, are dev and prd in same network segment? (just to rule that out - I'm wondering if ddboost connection between SQL server and DD gets broken).  Do you use multihomed setup?  If so, is the setup for sn/backup server interface identical for both prd and dev clients?

AdrianSahota1 · Answer

Hi Hrvoje, thanks for responding. I've tried Domain User & Local DBA user accounts and still no success.

Whilst trying to identify the problem, I've subsequently built a Development SQL AoAG Cluster, with the same Windows 2012 R2 Patch revision, and the same SQL 2012 SP1 patch revision, same users accounts and same privileges granted, and identical Networker configuration (Same group, same pool, same target device etc etc).

What i have found is, on the new Development Cluster, I can backup successfully, AND restore successfully.

On my Production Servers, if i select the Development Server as the source, I am also able to Restore the DEV Database to my production server (be it on AFTD or DDBOOST devices).

However, if i try and recover my Production Databases onto the newly created Development Server, I get exactly the same response i did previously;

NSR critical 34 Error opening tape session for %s. 1 0 16 MSSQL:/test_recv

NSR warning 34 VDI device did not close properly. 0

My Logs still aren't showing anything more than they were previously, I've compared successful to failed and there's no discernible difference to me as to why I cannot restore Production, but can Development!

ble1 · Answer

That's really strange.. you may wish to place NSR daemons on both ends into debug mode to see if that will get you more info.

AdrianSahota1 · Answer

Yes, Prod and Dev are in same network segment, and both have full accessibility to the DataDomain (and also our CIFS NAS server which i've also tried)

No to Multihomed, single NIC interfaces on all, all configured identically.

NetWorker

SQL 2012 Always On Recovery Failure

Was this post helpful?