Hello. I am using DPM 2010 on Windows 2008R2, and a PowerVault 124T autoloader. Recently DPM stopped writing to tape giving this error:

DPM encountered a critical error while performing an I/O operation on the tape VM-Weekly-00000028 (12-150L4) (Barcode - 12-150L4) in Drive IBM ULTRIUM-HH4 SCSI Sequential Device. (ID 30101 Details: The request could not be performed because of an I/O device error (0x8007045D))

After working with Dell, they saw that the autolaoder itself was generating errors and sent out a refurbished one. The refurbished one is now in place, with current firmware and drivers and the same configuration as the old one, however the backups to tape continue to fail with the above error message. The autoloader no longer generates error messages though.

I did find another post that suggested adding a registry key:

[HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Microsoft Data Protection Manager\Agent]
   "BufferQueueSize"=dword:00000002

which did nothing. The backup job still failed after transferring 13 Mb.

The nightly backup to disk works, and the tape drive shows up correctly in device manager. A driver check there says the correct driver is installed.

Any help is greatly appreciated!

• Edited by lavinrc 21 hours 49 minutes ago

What build level is your DPM 2010 installation at? Is the host 2008 R2 OS patched to SP1? Can you check your windows update history to see if any patch was applied intentionally or unknowingly prior to when you started to receive the error? Has either device been fully restarted since the hardware repair? If the tape library isn't functioning with DPM, can the IBM diagnostic tools be run to test connectivity and functionality?

Hi Kyle. Sorry, I was out for a couple of weeks. To answer your questions:

What build level is your DPM 2010 installation at? I'm not sure how to tell. Going to Hellp-->About... just says 'Version: ' (no number)

Is the host 2008 R2 OS patched to SP1? yes

Can you check your windows update history to see if any patch was applied intentionally or unknowingly prior to when you started to receive the error? 3/4 - updates applied --> every sunday until 5/5 successful tape backups --> 5/12 tape backup failed --> 5/14 more updates applied --> every sunday since 5/14 tape backups failed

Has either device been fully restarted since the hardware repair? In working with Dell, the tape library, SAS cable and SAS card have all been replaced. Both devices have been restarted multiple times. The library shows up in Device Manager and in DPM (although the drive seesm to disappear every once and a while)

If the tape library isn't functioning with DPM, can the IBM diagnostic tools be run to test connectivity and functionality? Dell has tried to run ITDT many times but it cannot connect to the library. It says there is something with a handle open but they cannot find any service (we stopped all of the DPM services) or handle (running handler) that is the culprit. Do you have any suggestions of what might need to be disabled for ITDT to run?

Futher - DPM successfully scans the library. I removed all of the tapes and replaced them (some of which have not expired yet). DPM inventoried the library correctly and listed the tapes ready for off site as well as the expired and free ones. It seems that DPM can in fact communicate again, it just can't write anything.

• Edited by lavinrc Wednesday, June 05, 2013 4:50 PM spelling

An update. After spending 6 hours on the phone with Dell today, they confirmed that the firmware version on the tape drive was too new for DPM and backed it down to the newest reported version that Microsoft has verified, removed all links to the tape library that they could, re-installed everything, made sure all appropriate Microsoft updates were installed. failures continue with the same error listed above.

• Marked as answer by lavinrc 21 hours 49 minutes ago
• Unmarked as answer by lavinrc 21 hours 49 minutes ago

A final update to clear out this question. Dell essentially replaced *every* hardware part they could and reloaded *everything*. Now it works 90% of the time.

• Marked as answer by lavinrc 21 hours 48 minutes ago