🚀 go-pugleaf

In article <JTKOHL.90Nov9083139@quicksilver.MIT.EDU>, jtkohl@MIT.EDU (John T Kohl) writes:
> I have a DS3100 on my desk, with the following normal configuration:
> 	Ultrix/RISC 3.1
> 	16M memory
> 	SCSI #0: external RZ55
> 	SCSI #3: internal RZ23 (root partition)
> 	SCSI #4: external RZ55
> 
> Two days ago I hooked up an external SCSI TK50 drive, at SCSI #5.  After
> doing so, my root partition was immediately trashed.  Removing the tape
> drive allowed me to reconstruct the partition; I then put the tape drive
> back on the system, and the drive was trashed again.

	I had this problem a long time ago.  It turned out it was
	broken SCSI electronics in the TK50.  I got the TK50 and
	haven't had a problem since.

	Check the error log to see if any hardware errors are
	getting logged.  I tended to get SCSI parity errors.
> 
> One co-worker here reports similar problems with an Exabyte drive
> interfering with his internal RZ23, and that when he accessed said
> RZ23, the Exabyte would occasionally make noises as if it were being
> accessed.

	Trouble shooting this sort of thing is easy.  Back up
	everything to wherever you can (network for example).
	Then use the disk exerciser in /usr/field to put a
	load on the system.  You can run it readonly if you
	don't have any partitions to scribble on.  Depending
	where you start add or remove devices to/from the bus
	until you find the one that causes the other to die.

	That is probably the broken one.
	up
> 
> Anybody else heard of such a problem/know a solution?

	Yup.
> --
> John Kohl <jtkohl@ATHENA.MIT.EDU> or <jtkohl@MIT.EDU>
> Digital Equipment Corporation/Project Athena


-- 
Alan Rollow				alan@nabeth.enet.dec.com