Dave wrote:

> Frank-Christian Kruegel wrote:
>
>> On Sun, 06 Nov 2005 15:53:04 +0000, Dave <nospam@nowhere.com> wrote:
>>
>>
>>> I have a 9 GB (hence reasonably old) disk which refuses to be seen
>>> even at the OK prompt with a probe-scsi in a SPARC 20. One might
>>> conclude the disk is dead, but it seems to work fine in a Ultra 80.
>>> Both machines run Solaris 9, although I patched the SS20 last night,
>>> so it is more up to date.
>>
>>
>>
>> The sun4m esp fast scsi controller is 8 bit only. I'm unsure whether the
>> upper 8 bits of the wide (16 bit) sca connector are terminated
>> properly or
>> have been left open. Some wide hard disks hang unless both lower and
>> upper 8
>> bits are properly terminated.
>>
>> Maybe jumpering the drive to single ended may help. Jumpers for drive
>> id or
>> termination have to be removed.
>
>
> Thanks for that.
>
> I checked and found the disk was already jumped SE. I removed the jumper
> (what else to try?) but no luck. I tried a few other jumpers, but no
> improvement.
>
> What was odd though is that
>
> a) The disk is seen if it is the *only* disk in the machine.
>
> b) The disk is seen if it the lower slot (SCSI ID=3), but not if its in
> the upper one (ID=1). And no, there are not any jumpers installed on
> any of the SCSI ID select pins.
>
> Anyway, it is at least semi-resolved. I've put an operating system on it
> and it boots OK now. Just needs to be in the lower slot, otherwise it
> can't be seen on the bus.

That's rather odd. There is no physical or logical difference between
the upper and lower drive bays aside from SCSI ID AFAIK.

I suspect your upper bay is b0rked somehow, and won't work with any disk
as a result.

I have a couple of spare SS20 backplane & cable assemblies here
(Toronto, Canada), perhaps you could try one if the logistics aren't too
complicated.

Triffid