Summary addendum: Using Sun 3/160 SCSI board for real SCSI drive?

From: Rob Scott (zeke@mpl.UCSD.EDU)
Date: Wed Jan 13 1993 - 12:07:05 CST


Greetings,

After I posted my initial summary about using an embedded SCSI drive on
a Sun 3/160 with the internal SCSI board, I received an additional reply
from Stefan Mochnacki <stefan@centaur.astro.utoronto.ca> that probably
is the solution to at least my particular case. I felt it important
enough to post this addendum.

Briefly, my initial post:
>> After removing the factory installed ST-506 disks, tape drive, and their
>> controller boards from a very old Sun 3/160, I installed a Micropolis
>> 1578 SCSI drive attached to the internal SCSI cable.
>>
>> When I try and access the drive, I can get to only parts of it successfully.
>> Accessing the drive always yields a large amount of message like:
>> "sd0: Excessive soft errors! Check drive."
>>
>> Some partitions can only be partially accessed, and attempts to access
>> them meets with failure and the message:
>> "sd0h error: sense key(0x3): media error, error code(0x11): hard data error"
>> "sd0h: read retry, block 531937 (142912 relative)"
>> for various sectors.
>>
My initial summary:
>> The general concensus was that I should have no problems using a standard
>> SCSI disk with the internal Sun SCSI interface card. There are several
>> caveats, however, and they are listed here in order of importance:
>>
>> 1) most Sun-2/Sun-3 SCSI cards don't supply termination power to the
>> TRMPWR line on the SCSI bus. Only very late engineering rev
>> cards supply power (if your's does it will have a fuse on
>> the card). Make sure that your disk is supplying termination
>> power (most disks have a jumper to enable/disable this).
>> 2) Make sure that there isn't double termination if you leave the
>> tape drive in and working. The tape drive controller card
>> is usually where the termination is in these older systems.
>> 3) Sun/OS 3.5 probably won't work with a real SCSI drive attached
>> to the SCSI card. This wasn't a problem for me, since I'm
>> running 4.1.1.

Stefan sent me several suggestions (all but one covered above), but included
one that I had neglected in my troubleshooting:

    The ROM in the 3/160 may need to be upgraded (means a new motherboard
    is you are on a maintenance contract).

This is likely my particular problem, since I believe that my VME cage
serial number is #31, and to my knowledge the CPU card has never had to
be exchanged because of failure.

                    -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
     Rob Scott, System Administrator for the Marine Physical Laboratory
            Scripps Institution of Oceanography, UC San Diego
    Internet: zeke@mpl.ucsd.edu Cserve: 76176,637 Voice: (619)534-5503



This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:07:21 CDT