MachineGreen:
Well the thread about tape spanning issues is an old one now. More recently I've been getting a weird problem where the SCSI bus resets itself and causes lockups.
My log files show errors such as "unit scsi0/etc..etc failed to respond in the timeout period" and "...detected a controller error" type messages. The thing is - the problem escalated over the period of a week - on multiple OSs (WinXP,Win98,Linux), and then inexplicably just stopped. For two days now I've had no problems, and no SCSI related error messages in the logs.
The only "change" to my system was I got infected by the W32.Welchia.Worm - it somehow got through my defences - but one clean install later I've no problems. What's weird about that, is that the Welchia worm does not affect Win98 or Linux at all, yet removing it seemed to solve my SCSI problem *even under those OSs*?
Another "change" (now I think about it) is that the two intake fans at the front of the case (the ones that blow right over the hard drives) were switched off for ... well weeks I suppose. I'd switched them off because I had a headache, and just plain forgot to switch them back on. Switching the fans back on, coincided with my "clean install" (I did a full hardware/software cleanup) - so maybe it was because they were overheating. They are Seagate Cheetah X15 (15K/rpm) U160 drives. Then again, all that fiddling inside the case maybe just nudged a loose cable or card back into place. Dunno. Even that backup (that I mentioned above) now restores without problem.
I need a live-in tech support department
Back to your problem:
Your assertion that “it worked before, it should still work” does not take into consideration the fact that cables and cards come loose through vibration and heat expansion/contraction. Also all hardware packs up eventually. Then there’s Viral infection and software/resource conflicts (especially on Microsoft systems where the OS quite happily makes major system changes in the background without even telling you). On more than one occasion, I’ve discovered that the sole reason for unexplained computer failure was … dust – which is an insulator. Too much dust (especially directly on chips – E.g. memory chips) can cause overheating to the point of lockups and blue screens.
Just because a SCSI card “seems” to be working with other devices, does not mean it is 100% functional. Maybe those other devices don’t stress the I/O as much as your tape streamer. Maybe the problem is related to a background service (daemon) interfering with the backup software (an earlier update to Norton Anti-Virus seemed to have that effect on Novanet).
The news that you had earlier problems due to hot weather leads me to believe that the SCSI card has blown something minor, which only manifests itself whilst operating the tape drive – but that’s just a gut feeling. You need to isolate the problem. The fact that multiple devices all connected to the same card (correct?) all start failing – looks suspicious to me. If I’d lost 3 drives and a tape streamer in a single blow, I’d be suing the SCSI card manufacturer.
[H]omer