Re: A single SCSI disk scanned multiple times
- From: dnichols@xxxxxxxxxxx (DoN. Nichols)
- Date: Wed, 22 Feb 2006 06:14:31 +0000
According to <wcui@xxxxxxxxxx>:
Hi,
We have a Solaris machine with a 36GB Seagate ST336607LC SCSI boot
disk. Recently it was noticed that when the system boot up, the same
boot disk will be scanned 16 times! Of course, there is only one boot
disk, which name is c1t5d0, the other disks are named c1t1d0, c1t2d0,
c1t3d0, ...c1t15d0. These disks can be observed by the 'prtconf' and
'format' program. But the tape drive on the same SCSI bus wasn't
observed. When we didn't plug the boot disk, the tape drive will be
seen, and the bogus disks will not appear. But when we plug the boot
disk, the bogus disks will appear and the tape drive will disappear. We
tried to replace the SCSI cable and SCSI terminator, but no progress.
Have you exprience on this strange problem? Any suggestion is
appreciated.
It sounds to me as though the connection to the SCSI ID jumpers
is wrong. Looking at the Seagate web page, I find that it does have a
set of SCSI ID jumpers on the front of the drive. The upper row
(starting with pin 1 (closest to the edge)) are the address pins,
starting with A0, then A1, then A2, and finally A3 (followed by another
pair of pins for another function labeled "RES", and a pair for an
external activity LED to connect to.) The bottom row of pins (at least
for the address pins) are all ground.
There is another set of jumpers beside the 68-pin connector, but
I think that one of these jumper areas is the source of your problem.
Anyway -- sometimes the SCSI id connector is put on upside down.
Normally, it connects one or more pins to ground to select specific SCSI
IDs. And sometimes only one ground connection is made to a single pin
on the connector, but other times, it is made to all four of the ground
pins.
Now -- if the connector is put in upside down, it connects all
of the *address* pins together, and can result in the drive responding
to all sixteen addresses (and masking your tape drive as well). (Or all
eight addresses for a 50-pin SCSI, which you don't have in this case).
So -- examine the drive, and look for the connector setting the
SCSI addresses. Unplug it and turn it over before re-connecting it.
You don't specify which Sun system (just "Solaris machine",
which could be any of a large number of hardware platforms), and you
don't say whether the drive is internal or external. I have found this
kind of problem more often with external drives, as internal ones either
will have individual address jumpers on the pins without a cable, or
will have an SCA interface (80-pin, which includes address selection
jumpers and power cables in the single connector) and the backplane into
which the drive is plugged will select the address for the drive.
But -- there may be Some Sun platform which has an internal
cable selecting SCSI ID on the drive.
If you want to read up on the drive and the function of the
connectors, you can download the manual from Seagate's web site. Check
out:
http://www.seagate.com/
and put your drive model number into the search by model number blank.
You can then download the manual (a 96-page PDF file) which will tell
you more than you ever wanted to know about the drive.
Good Luck,
DoN.
--
Email: <dnichols@xxxxxxxxxxx> | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---
.
- Follow-Ups:
- References:
- A single SCSI disk scanned multiple times
- From: wcui
- A single SCSI disk scanned multiple times
- Prev by Date: Re: Sun Microsystems External Mammoth Tape Drive
- Next by Date: Internal DDS3 causing SCSI bus issues on SB1K
- Previous by thread: Re: A single SCSI disk scanned multiple times
- Next by thread: Re: A single SCSI disk scanned multiple times
- Index(es):
Relevant Pages
|
