V-79-57344-34029 - A hardware error occurred

[pugsley42]

Setup:

Server - HP DL380 G4

OS - Windows Server 2003, SP1

Backup hardware - HP StorageWorks SDLT320

Backup software - Symantec Backup Exec 12.5, SP4

We have been having issues running backups on this server for quite awhile, and keep getting this error (among a couple of others).  Each time it wil backup for what seems a random amount of time, then fail with this error.

What we've tried:

Replaced the SDLT320 with a known good working unit

Replaced the SCSI card with a brand new one

Replaced the SCSI cable with a brand new one

Reinstalled BE 12.5

Updated BE 12.5 to SP4

Service account is a Domain Admin account

The latest attempt was a full backup that backed up about 16GB of data in 229min.  We only watched for the first 30min but saw many files being backed up.  Looking at the logs in the morning it errored out on almost every directory of the RAID set saying the file or directory does not exist.  When going through to perform a restore to see what data was backed up it only showed AD information.  Following this, it performed an incremental where we received the same error as in previous attempts, about the hardware error.

Any suggestions?

[CraigV]

have you checked to see if your RAID set is healthy? I have had an issue where a drive that failed caused my backups to fail...the remaining drives struggled that much!

[Mahesh_Roja]

Refer this KB 

How to troubleshoot issues with a Robotic Library (autoloader/changer) and/or Tape Drive(s)

http://www.symantec.com/business/support/index?page=content&id=TECH24414

Troubleshooting hardware with the Backup Exec for Windows Servers using the SCSI Trace Utility (tracer.exe). 
 http://support.veritas.com/docs/285630 
 

[pugsley42]

@CraigV - I ran a chkdsk /f on the RAID set and it came back fine.

@Maheshroja - It's a single tape drive, not an autoloader.  I have not run the SCSI Trace Utility yet, and the tape drive is set to SCSI ID 5.  I'll have to look again at the Event ID's, but they were all 5 digits.

[Mahesh_Roja]

Its resolved the issue means mark as solution which has helped

[pugsley42]

It's not resolved yet.  Nothing has changed.  Next, I'll be trying to update the server to SP2 and testing via HP StorageWorks Library and Tape Tools.

[pugsley42]

SP2 for Server Win2k3 did not work.  Trying the updated StorePort drivers and the HP StorageWorks Library and Tape Tools test right now.

[pugsley42]

The Library and Tape Tools came back positive.  No problems.  Installing the StorePort drivers didn't help either.  I'm trying Backup to Disk right now, to an external 1TB USB HD.

[CraigV]

Hey man,

Couple of other things to try...

1. Upgrade the firmware on the SDLT through HP LTT.

2. Upgrade the Symantec drivers for BE 12.5, and then use the Device Configuration Wizard to install them for the SDLT.

3. Disable your AV and see if this works. If it does, explicitly exclude the BE service *.exes from scanning...

Craig

[pugsley42]

I have tried disabling McAfee, especially the On-Access scan, but that hasn't helped.  We have tried doing a straight file copy to an external USB drive just using Windows Explorer, but that will fail and give an error about the path or filename being too long on some entry.  Can I presume that wouldn't be an issue when using BE?  Last night before I left we set it up to do a Backup to Disk to the USB drive.  I'll have to check that later when the lab unlocks.

Where would I get updated Symantec drivers?  I must have missed that somewhere.  I haven't tried updating the firmware on the SDLT yet.  I tried another SDLT320 from another lab that works, and just presumed that if the firmware was sufficient on the device to work in the other lab, it wouldn't be an issue in this one.

[CraigV]

That error is a Windows error...it has nothing to do with BE.

You can find the latest drivers for BE 12.5 here:

http://www.symantec.com/business/support/index?page=content&cat=DRIVERS&key=15047&basecat=DOWNLOADS&...

[pugsley42]

It appears the Backup to Disk failed.  That takes the SCSI card (and bus), cable, tape and tape drive out of the equation, leaving configuration of the server or likely the server motherboard, which is 5yrs old and out of warranty.  Unfortunately the user didn't tell me the error message and I haven't been back to check the logs.  At this point, we already have a new server and autoloader waiting in the wings but scheduled to replace at a later date.  I think the time table just got moved up.

I appreciate all the help.