Video Screencast Help
Symantec to Separate Into Two Focused, Industry-Leading Technology Companies. Learn more.

SQL Restore Problem

Created: 13 Jun 2012 | 3 comments

So here is an unusual problem. I have reached out to Symantec Support, but they have yet to figure out whats going on. So we are performing a Restore of a SAP Database, from Tape, to an iSCSI target on another server. We ran it last week and it worked fine. So I dont know if a setting got changed.

So the Database is about 1.2TB in size. When we run the restore, everything goes as normal. During the restore, if you check the destination, it shows 1 MDF, 12 NDF's, and 1 LDF. Total size reported is about 1.2TB. After the restore completes, which takes about 4 hours, the data changes. Up until 99%. the correct data is showing on the destination. But, once it hits 100%, the files are replaced with 1 new MDF and 1 new LDF, totalling about 140MB.

We are using the SQL Redirect option, with Overwrite Database enabled. Symantec has looked at it, but doesnt see anything wrong yet. The job completes, and even reports a Success, without any warnings or errors. I think it must be some setting somewhere, but I cant figure out where.

Any suggestions would be great, Thanks!

Here is the job summary:

Restore Job Properties: SAP CCP DB RESTORE

Media Labels

Media required for restore job: 000056L3

Source

Selections

Selection list name: SAP CCP DB RESTORE

Resource Credentials

Resources Resource Credentials
ndcsapccop02.PCCR.US System Logon Account
Microsoft SQL Server  

Destination

File Redirection

Redirect file sets: No

Microsoft SQL Redirection

Redirect Microsoft SQL Server sets: Yes
Restore to server: NDCSAPCCOQ01
Server logon account: System Logon Account
SQL logon account:  
Restore to named instance:  
Restore to database: CCPTESTNEW
Use this path: E:\CCPRESTORE
Retain replication information: No

Settings

General

Job name: SAP CCP DB RESTORE
Job priority: Medium
Restoring existing files: Overwrite the file on disk only if it is older
Restoring security information: Restore all information for files and directories
Restore corrupt files: No
Preserve tree: Yes

Advanced

When restoring registry information: Merge the existing hardware configuration and registry services with the data to be restored
Junction points: Restore junction points, symbolic links, files and directories from backup media
Restore Removable Storage data: No
Restore disk quota data: No
Restore Terminal Services database: No
Restore Windows Management Instrumentation repository: No
Restore cluster quorum: No
Mark this server as the primary arbitrator for replication when restoring folders managed by the File Replication Service, or when restoring SYSVOL in System State: No
Restore Windows NT 4.0 registry: No
Path on media server for staging temporary restore data when restoring individual items from tape: C:\TEMP

Microsoft SQL

Recovery completion state: Leave database ready to use. Additional transaction logs cannot be restored
Consistency check after restore: Full check, excluding indexes
Take existing target database offline: No
Overwrite the existing database: Yes
Automate master database restore: No
Continue restoring if an error occurs during the restore: No
Run verify only; do not restore data: No
Recover the entire log: Yes

Notification

Include job log with a notification to an e-mail or printer recipient: No

Frequency

Schedule

Run now
Submit job on hold: No
Enable automatic cancellation: No
Job deletion: Do not delete the job

Comments 3 CommentsJump to latest comment

Carlos_Quiroga's picture

I think I was sitting next to the TSE thet worked with you on your case...

Did you try to choose the "default instance disk location" on the redirection?

Regards,

TedThrive's picture

She had suggested that, but based on the size of the database, we cannot restore to the default location. The server which has SQL doesnt have 1.2TB of available space, which is why we are using the target location. I looked, and if there was any way that we could fit it, I would try that, but its just not feasible. Thank you for the suggestion though.

Carlos_Quiroga's picture

I would suggest then to create a new instance, and configure that instance to store the files in a place where there is space.

Redirect the restore to that instance, choosing the default path for that instance as the destination.

Regards,