Hi,

I'm trying to install a Seagate Barracuda ST318436LC SCA drive on my Sparc
5.

The problem is that the drive never spins up. I've tried many combinations
of the jumpers on board the disk, tried plugging it on the upper and lower
SCA slot, booted a gentoo netboot image (which tries to spin it up) but
everything fails.

I've tried the FORCE SE, MOTOR START ENABLE, DISABLE PARITY jumpers, no
change.

Now, the disk gets detected correctly by gentoo (it's led also blinks), but
probe scsi lists the drive on all 7 scsi ID's (plus when it does so, the
screen flashes and only the (c) Seagate message is visible), but I can see
it lists the drive on all scsi ID's. Gentoo detects it on ID 3 (lower SCA
connector).

Anyway, the disk itself doesnt have any termination jumpers, so, how is the
drive terminated? Does the sparc provide termination?

Maybe it's a OBP problem? Version on my sparc is 2.15, maybe a newer one is
needed to drive this disk? Or should I just consider the disk dead?

I've seen on the net reports that LC drives (the ones without termination)
do work on SS5's...

Unfortunatelly I dont have a way to test the disk on another machine.

Any help?

Thanks,
Emmanuel
--
If something is hard to do, it will happen.
If something is impossible to do, it'll just take a bit longer...

MrZammler <none@nowhere.com> writes:
>I'm trying to install a Seagate Barracuda ST318436LC SCA drive on my Sparc
>5.

I had that drive working in an Ultra 1 no problem once. Not in a Sparc
5, but other drives in the 5s worked fine.

>The problem is that the drive never spins up. I've tried many combinations
>of the jumpers on board the disk, tried plugging it on the upper and lower
>SCA slot, booted a gentoo netboot image (which tries to spin it up) but
>everything fails.

Sounds like a bad drive. Its pretty much a no brainer to install an
SCA drive into Sun's of this generation.

>Anyway, the disk itself doesnt have any termination jumpers, so, how is the
>drive terminated? Does the sparc provide termination?

Yes, the drive backplane on any SCA system terminates the bus appropriately.
You don't have to worry about termination. Same thing with ID. You
mention seeing it on ID 3, but the SCA backplanes used ID 0 as the
boot drive. You shouldn't be jumpering the SCSI ID either on this
drive. Just leave off any ID jumpers, and plug it in.

>I've seen on the net reports that LC drives (the ones without termination)
>do work on SS5's...

Yes, I've put in Seagate ST*LC drives into SS5s and Ultra 1s and 2s.

According to MrZammler <none@nowhere.com>:
> Hi,
>
> I'm trying to install a Seagate Barracuda ST318436LC SCA drive on my Sparc
> 5.
>
> The problem is that the drive never spins up. I've tried many combinations
> of the jumpers on board the disk, tried plugging it on the upper and lower
> SCA slot, booted a gentoo netboot image (which tries to spin it up) but
> everything fails.
>
> I've tried the FORCE SE, MOTOR START ENABLE, DISABLE PARITY jumpers, no
> change.

You shouldn't need to change *any* jumpers on a SCA drive
connected to a Sun -- or anything with an SCA backplane.

The upper slot on the SS-5 is SCSI ID 1, and the lower slot is
SCSI-ID 3. The system normally boots from SCSI-ID 3 on that generation
of machines -- and the ones back to the SS-1, which are *not* SCA
interface), and uses SCSI-ID 1 as the second disk. There was some claim
that access to the disk would be a bit faster at the higher SCSI ID.

However, with the Ultra-1 and later, they went back to booting
from SCSI-ID 0.

> Now, the disk gets detected correctly by gentoo (it's led also blinks), but
> probe scsi lists the drive on all 7 scsi ID's (plus when it does so, the
> screen flashes and only the (c) Seagate message is visible),

That blinking is because that particular drive (actually,
several of the Seagate 18GB drives) happens to include the ANSI sequence
for "home cursor, clear screen" in its ID message -- for lord only knows
what reason. Later machines, such as the Ultra-1 and later, include in
the OBP the option:

ansi-terminal?=true

which if you change it to read false, will probably not clear the screen
on receipt of those characters. But -- checking the two SS-5 machines
which I have running at present, that is not included in that version of
the OBP.

> but I can see
> it lists the drive on all scsi ID's.

Hmm ... do you have a CD-ROM connected externally? Normally,
the internal CD-ROM is jumpered as SCSI-ID 6, but the external usually
has a connector from a switch at the back of the drive housing (usually
a sandwichbox), and if that connector is upside down (it is only a 6-pin
connector which goes where the jumper blocks otherwise would go) and if
some SCSI ID other than 0 is selected (it should be 6), it has the
problem of tying all of the SCSI ID lines together and to ground, so
that drive responds to *all* addresses. (If you have been playing with
SCSI-ID jumpers on the SCA drive (the pins are there, though they should
be empty under normal circumstances), it may be that somehow the SCSI ID
lines are shorted together on the drive.

> Gentoo detects it on ID 3 (lower SCA
> connector).

That probably is because the EEPROM settings tell it to check
there first:

boot-device=disk

tells it (on the SS-5) to try to boot from SCSI-ID 3.

You can see all of the EEPROM settings by typing "printenv" at
the OBP level -- no need to boot.

> Anyway, the disk itself doesnt have any termination jumpers, so, how is the
> drive terminated? Does the sparc provide termination?

Yes.

> Maybe it's a OBP problem? Version on my sparc is 2.15, maybe a newer one is
> needed to drive this disk? Or should I just consider the disk dead?

I think that the disk has physical problems -- it won't spin up,
but the logic card identifies it anyway. I've seen others with similar
problems. You *might* be able to break it loose to spin up by bumping
it and then powering it up -- but it will just stick again soon enough
so it is not worth the trouble.

> I've seen on the net reports that LC drives (the ones without termination)
> do work on SS5's...
>
> Unfortunatelly I dont have a way to test the disk on another machine.
>
> Any help?

Get another disk -- these days they go for often less than the
shipping charge on eBay auctions.

An alternative OS which I have used on the SS-2 is OpenBSD --
which is about to release version 4.0. It is a good choice if security
is your most important criterion.

Good Luck,
DoN.
--
Email: <dnichols@d-and-d.com> | 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 ---

DoN. Nichols wrote:

>> it lists the drive on all scsi ID's.
>
> Hmm ... do you have a CD-ROM connected externally? Normally,
> the internal CD-ROM is jumpered as SCSI-ID 6, but the external usually
> has a connector from a switch at the back of the drive housing (usually
> a sandwichbox), and if that connector is upside down (it is only a 6-pin
> connector which goes where the jumper blocks otherwise would go) and if
> some SCSI ID other than 0 is selected (it should be 6), it has the
> problem of tying all of the SCSI ID lines together and to ground, so
> that drive responds to *all* addresses. (If you have been playing with
> SCSI-ID jumpers on the SCA drive (the pins are there, though they should
> be empty under normal circumstances), it may be that somehow the SCSI ID
> lines are shorted together on the drive.
>

Thanks for your answers, it cleared a lot of things up.

I did have a cdrom connected externally, but I dont now. Anyway, the listing
on all scsi id's (the sca drive is the only scsi thing connected on the
SS5) makes me wonder if the sparc has a general scsi problem or
something... ? Maybe that's why the spin up command doesnt work...

Although on the other hand, I've seen similar reports on google (a disk
listed on all scsi id's) but it shouldnt really be a problem (i.e.
eventually the disk should work).

I'm gonna mess around it some more today, but I guess it's almost time to
pronounce the disk dead.

Thanks all,
Emmanuel
--
If something is hard to do, it will happen.
If something is impossible to do, it'll just take a bit longer...

In article <1161674605.953316@athprx03>, MrZammler <none@nowhere.com> wrote:
>DoN. Nichols wrote:
>
>>> it lists the drive on all scsi ID's.
>>
>> Hmm ... do you have a CD-ROM connected externally? Normally,
>> the internal CD-ROM is jumpered as SCSI-ID 6, but the external usually
>> has a connector from a switch at the back of the drive housing (usually
>> a sandwichbox), and if that connector is upside down (it is only a 6-pin
>> connector which goes where the jumper blocks otherwise would go) and if
>> some SCSI ID other than 0 is selected (it should be 6), it has the
>> problem of tying all of the SCSI ID lines together and to ground, so
>> that drive responds to *all* addresses. (If you have been playing with
>> SCSI-ID jumpers on the SCA drive (the pins are there, though they should
>> be empty under normal circumstances), it may be that somehow the SCSI ID
>> lines are shorted together on the drive.
>>
>
>Thanks for your answers, it cleared a lot of things up.
>
>I did have a cdrom connected externally, but I dont now. Anyway, the listing
>on all scsi id's (the sca drive is the only scsi thing connected on the
>SS5) makes me wonder if the sparc has a general scsi problem or
>something... ? Maybe that's why the spin up command doesnt work...
>
>Although on the other hand, I've seen similar reports on google (a disk
>listed on all scsi id's) but it shouldnt really be a problem (i.e.
>eventually the disk should work).

If a disk responds on all SCSI IDs it is jumpered to the same ID as the
Host Adapter (controller). ID=7. Or the disk is badly broken.

carl
--
carl lowenstein marine physical lab u.c. san diego
clowenst@ucsd.edu