Backup Exec 10 communication problem with wide SCSI DLT1

[Robert_Becker]

My server is a quad processor Dell with RAID 5 and includes an
Adaptec 7899 (chip on the motherboard) 68-pin external SCSI port
(equivalent of a U160 controller)
Adaptec 2940U2W, also with 68-pin external SCSI port (Ultra 2 Wide)

I have two identical tape drives, both Quantum Benchmark DLT1's (one is Dell oem'd). The Dell oem'd drive is on the Adaptec 7899 and the Quantum Benchmark is on the Adaptec 2940U2W. The Quantum Benchmark is brand new; the Dell oem'd Benchmark is 5 years old.

I am getting the following error on a regular basis on both tape drives:
----------------------
Data error (cyclic redundancy check).
This could be caused by either a dirty tape drive, bad media, or a SCSI problem.
V-79-57344-34028 - Error reading/writing data from/to the media.
----------------------

I have followed the directions listed by looking up that error code. I have cleaned the heads repeatedly, the SCSI cables are in excellent shape, connections are solid. Both drives have active termination attached and the lights are on. Both drives are the only devices on their SCSI chain, and both are SCSI ID 2.

I've followed the instructions in http://support.veritas.com/docs/191158 (How to resolve SCSI bus timeouts) without success.

I must be missing _something_. Help?

Robert

[Robert_Becker]

I've discovered the Adaptec drivers and firmware for both controllers are either not installed correctly or old. I'm reinstalling them this evening.

[Robert_Becker]

Correct drivers are now loaded for all non-RAID SCSI controllers. Last night's backup on the Dell DLT went 56 GB before failing with the same error.

The system has all of Microsoft's latest security updates.

The version of Backup Exec, exactly, is
10.0 Rev. 5484

As far as I can tell it has none of the BackupExec service packs installed.

[Robert_Becker]

Latest results: three of the last four Exchange backups (which are 27 GB) have worked. The last two full backups (72 or so GB) which run to two tapes have worked. I'm going to run the system as is for the next week and examine the results. Also, for diagnostic purposes I am going to label each tape that backs up successfully as "good" and the ones that don't as "bad" and see how that affects the percentage of successful backups. Assuming the software and OS is working together successfully, this percentage of successful backups should increase. If the software and OS is not working together successfully, it will not change the percentage of successful backups.

Wish me luck.