Error w/Autoloader, backups fail because of hardware issue

[kcoect]

I currently have an HP 1/8 G2 Ultrium 2 Autoloader that has been working for years without a problem, but recently I have been having my backups fail with the following error:

Backup- C:
Storage device "HP 2" reported an error on a request to write data to media.
Error reported:
The request could not be performed because of an I/O device error.
V-79-57344-34029 - A hardware error occurred.

Now, I've make sure the loader and tape drive are up to date with the latest firmware, I've made sure I have the latest device drivers for BackupExec (v11d is what were running) as well as the latest patches for BackupExec. I've also doublechecked the cables and terminator to make sure there seated well.

This server is in a locked and climate controlled room, so nobody can get to it.

Any help would be greatly appreciated! Thanks!

[kcoect]

I forgot to mention that I've also run the HP StorageWorks Library and Tape Tools and run the diagnostics on the drive and autoloader. The autoloader passes fine, but the drive is reporting problems that are usually related to cables or SCSI termination. Other times when I run the test, they pass sucessfully. It doesn't make sense that the autoloader would all of a sudden start having these issues with cables or termination after it's been in use for over a year...

[kcoect]

I should have put all of this in one message, sorry for the seperate replies...

When I run the "Connectivity Test" I get the following error or failure:

LTO_SM_BufferRdWr:Check condition sense value returned: SenseKey = 0b,ASC/ASCQ = 47/01

Connectivity Test - FAILED.
Please check the configuration of your SCSI Bus, proper termination, proper installation of all required drivers. Then re-run the test.

[thomasr]

parallel SCSI cables do sometimes "just go bad" -- perhaps the bend is a bit too tight, or an internal connection was never too good. Consider powering the unit down, then reseating all cables, plus removing and re-inserting the drive. If you have cables you could swap even temporarily, it's worthwhile to see if a different cable makes a, er, difference.

Have any other devices been added to the SCSI device chain recently? In one case I remember from a while back, we'd simply exceeded the pSCSI supported cable length, and things became unpredictable, much as you mention above.

[kcoect]

I'm going to look up a replacement cable this morning. I know that will be one of the first things that support asks about and unfortunatly I didn't have a replacement when I was on-site yesterday.

I did power down completely both the server and tape drive, unplug power, remove and reseat all the SCSI cables, and then brought everything back up to no avail. Nothing has been added to the SCSI chain and the Autoloader is the only thing connected to that controller. Also, the SCSI chain is terminated at the tape drive too, I checked that as well...

Time to go find a cable...