Category Archives: Backup

ID 30101 Details: The request could not be performed because of an I/O device error (0x8007045D)

On System Center Data Protection Manager (DPM) 2010 with March 2011 hotfix I got the following error when backing up to tape:

ID 30101 Details: The request could not be performed because of an I/O device error (0x8007045D)

This is an error that was originally fixed in DPM 2007 with a hotfix http://support.microsoft.com/kb/970868 , but reappeared on my 2010 install.  It can occur with large tape backups that span more than one tape.  I fixed the error by adding the following registry key, a reboot wasn’t necessary:

[HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Microsoft Data Protection Manager\Agent]
"BufferQueueSize"=dword:00000001

[UPDATE]

I’ve had the same problem with DPM 2012 SP1 and the registry still fix worked.

Advertisements

System Center Data Protection Manager 2010 Doesn’t Detect Tape Drives In Library

[UPDATE – June 2013]

I’ve just encountered the same issue using Windows Server 2012, System Center Data Protection Manager 2012 SP1 and an Overland REO 4600 VTL.  As with DPM 2007, the fix is to change the tape drive driver to the generic Microsoft LTO tape driver.

I’ve used System Center Data Protection Manager (DPM) since the 2007 beta, primarily for SharePoint and SQL backups.  I have a primary DPM server which backs up to disk and a secondary server which backs up to disk and additionally to a Overland Storage REO 4500 iSCSI Virtual Tape Library (VTL) for long term protection.  Both servers have run Windows Server 2008 R2 with DPM 2010 since its release.

I’ve never had any major issues with DPM until I recently rebuilt the secondary server from scratch with a clean OS install of Windows Server 2008 R2 SP1 and DPM 2010.  What I found was that although DPM installed correctly it couldn’t see the tape drives in the VTL.  All the tape drives and the medium changer were visible in Device Manager and none were showing errors.  The medium changer and driver were automatically installed by Windows and the tape drive driver came from Windows Update.  The REO 4500 emulates an HP LTO2 tape drive and this is what was displayed in Device Manager.

In the DPM console under Management -> Libraries, the library was listed, but Total Drives was 0, rather than the expected 16.  The DPM console showed the following error:

Tape Library Overland Neo Series Medium Changers is not functioning efficiently for the following reason:

The drive information for Tape Library Overland Neo Series Medium Changers is incorrect and needs to be refreshed. (ID: 3302)

I followed the instructions to remap the drives here: http://technet.microsoft.com/en-us/library/bb795782.aspx.  The DPMLA.xml file produced by the tool was correct, showing the correct SCSI IDs, serial numbers etc. but DPM still couldn’t see the drives.

I found the solution to the problem on the DPM forum here: http://social.technet.microsoft.com/Forums/en-US/dpmtapebackuprecovery/thread/1d599443-7bf6-437a-bf12-52847fa7c8e5/ What I did is update the tape drive driver to the Inbox LTO driver as described on the forum and below:

1) Open device manager
2) Locate the tape drive
3) Right-click and look at the properties.
3) Under the DRIVER tab, select UPDATE DRIVER
a) Select the Install from list or specific location (Advanvced) – next.
b) Select Don’t search. I will choose the driver to install. – next.
c) Uncheck the Show comtabile hardware checkbox.
d) Highlight LTO under the manufacturer.
e) Highlight the LTO tape drive under model – then next.
f) This should install the Microsoft ltotape.sys driver.
4) Rescan the tape library in the DPM console – try to take another backup.

After following the above steps, all 16 tape drives were visible in the  DPM console.

Backup Exec and Autonomy IDOL

Autonomy IDOL enables you to scheduled a backup of its index to local disk.  I wanted to use Backup Exec to backup the Autonomy backup.  The Backup Exec agent was installed, but the service was refusing to start.  By default the Backup Exec agent service communicates on port 10,000, which was being used by IDOL.  Changing the Backup Exec agent port enabled the service to start.  See here for details: http://www.symantec.com/business/support/index?page=content&id=TECH24256 

The case of the failed SQL backup

I enabled a maintenance plan on a SQL server which creates a full backup of all databases.  When checking the job history I noticed that the job had failed.  Upon further investigation I could see that although the job had failed, all the databases had been backed up.  Opening the job log revealled the following error:

Failed:(-1073548784) Executing the query “BACKUP DATABASE [master] TO  DISK = N’I:\\Microsoft SQL Server\\MSSQL.1\\MSSQL\\Backup\\master\\master_backup_201010260000.bak’ WITH NOFORMAT, NOINIT,  NAME = N’master_backup_20101026000019′, SKIP, REWIND, NOUNLOAD,  STATS = 10
” failed with the following error: “Transaction (Process ID 580) was deadlocked on lock resources with another process and has been chosen as the deadlock victim. Rerun the transaction.
Could not insert a backup or restore history/detail record in the msdb database. This may indicate a problem with the msdb database. The backup/restore operation was still successful.

I checked the SQL job schedule and noticed that another “Cleanup History” maintenance plan was set to run at the same time.  In order for this maintenance plan to cleanup the job history is needs to access MSDB, which caused the deadlock.  Re-scheduling the Full Backup plan to run 5 minutes later cured the problem

Partition above extends beyond the end of disk

I was recently trying to backup a server using Symantec Backup Exec System Recovery (BESR) 2010.  Unfortunately, BESR was refusing to create a backup.  A support call to Symantec pointed us to the PartInfo.exe tool, located in C:\Program Files\Symantec\Backup Exec System Recovery\Utility.  Running this tool creates a text file called partinfo.txt in the same folder.  When PartInfo was run on this particular server it logged the error “Partition above extends beyond the end of disk”.  This machine was a virtual server running on Windows Server 2008 R2 and had converted from a physical machine.  I was able to resolve the error by shutting down the VM and increasing the size of the disk by 1GB (the smallest amount permitted by the System Center Virtual Machine Manager 2008 R2 GUI).  I restarted the VM, re-ran PartInfo and the error was resolved and BESR was able to take backups of the server.