Video Screencast Help

Unable to restore clustered sql database

Created: 14 Feb 2013 • Updated: 18 Feb 2013 | 4 comments
wv's picture
This issue has been solved. See solution.


i'm trying to restore clustered datbase to same node with different name (previous month) . It fails with insuficient disk space. despite we have 10times space available restore fails.

DB size is approx 103GB restore LUN has 1.5 TB free space

NBU  windows 2003 r2

SQL 2008 r2 clustered

Comments 4 CommentsJump to latest comment

V4's picture

Is drive where u are restoring part of cluster? If not then  

Try adding that drive in cluster resource group and then perform restore

also post dbclient logs 

Marianne's picture

Please post restore script as well as dbclient log on destination client.
Any chance you are accidently restoring to inactive node?

Supporting Storage Foundation and VCS on Unix and Windows as well as NetBackup on Unix and Windows
Handy NBU Links

RLeon's picture

There are a few requirements when restoring to a SQL cluster.

1. You must first (re)start the SQL server/services in single instance mode. This is some kind of maintenance mode. Basically you have to first un-cluster the cluster.

1a. Now you should only have a single active SQL server, let's call it NodeA. This is where you should run the SQL restore procedures. You cannot restore SQL from within the master or media servers. You have to do it from within the active SQL server NodeA.

2. Since you backed up the SQL cluster using a virtual name, NodeA cannot restore the DBs because Nbu treats the virtual name and NodeA as unrelated entities.
To solve this problem, one solution is to create the following touch file:

3. Now that the prerequisites are all set, you have to follow the steps in the Nbu SQL admin guide under "Redirecting a restore to a different host". It is considered a redirect restore because, as stated previously, the SQL virtual name and NodeA are considered unrelated by Netbackup.

You will find details to all 3 of the above steps in the Nbu SQL admin guide.

wv's picture

hello all

sorry for delay in responding. was stuck with back to back restore calls.

following things was done to complete restore. Thanks to capt.jack for hint

we verified we have No.Restrictions is place

hostname/virtualname used during restore was correct

drive space was sufficient

user had privileges required for restores

disk was also member of cluster group. but when we did failover due to patch update on nodes we found disk didn't failed over. so suspect was correct somehow disk was not correctly became part of cluster group

we removed it and added new LUN and did failover drill which went successful and subsequent out restore too.

Thanks jack for highlighting disk part