[comp.sys.amiga] strange hard disk behavior

bader+@andrew.cmu.edu (Miles Bader) (01/02/89)

Hi, lately, either my hard disk (q80s) or my controller (2090a) has
been acting very strangely: If I reboot, the reboot succeeds to the
point where the hard disk normally takes over (just after
binddrivers).  But at that point, the drive light will turn on and
stay on, with seemingly no disk activity (that I can hear), and the
system hangs at that point [I've waited up to 15 mins. for something
to happen].  If I reboot with C-A-A at this point, it boots just fine.
But if I get a guru, the cycle starts over again....

Anyone have any idea what's up?  This isn't too inconvenient, but I'm
worried it's evidence of impending doom...

Thanks, -Miles

grwalter@watcgl.waterloo.edu (Fred Walter) (01/04/89)

In article <IXjeaqy00UkaMkKllB@andrew.cmu.edu> bader+@andrew.cmu.edu (Miles Bader) writes:
>Hi, lately, either my hard disk (q80s) or my controller (2090a) has
>been acting very strangely: If I reboot, the reboot succeeds to the
>point where the hard disk normally takes over (just after
>binddrivers).  But at that point, the drive light will turn on and
>stay on, with seemingly no disk activity (that I can hear), and the
>system hangs at that point [I've waited up to 15 mins. for something
>to happen].

Geez, the same sort of thing happens to me every so often (2090A,
Miniscribe 8051S). I reboot and the machine freezes every so often (usually
in BindDrivers). If I try to move the mouse, well she don't move at all. If
I reboot again everything is fine.

I too am worried about this behaviour ... (my 2090A is under warrentee for
another 10 days... if this is a hardware problem I hope someone can tell me
about soon).

	fred

jim@b11.INGR.COM. (Jim Levie ) (01/06/89)

In article <7546@watcgl.waterloo.edu> grwalter@watcgl.waterloo.edu (Fred Walter) writes:
>In article <IXjeaqy00UkaMkKllB@andrew.cmu.edu> bader+@andrew.cmu.edu (Miles Bader) writes:
>>Hi, lately, either my hard disk (q80s) or my controller (2090a) has
>>been acting very strangely: If I reboot, the reboot succeeds to the
>>point where the hard disk normally takes over (just after
>>binddrivers).  But at that point, the drive light will turn on and
>>stay on, with seemingly no disk activity (that I can hear), and the
>>system hangs at that point [I've waited up to 15 mins. for something
>>to happen].
>
>Geez, the same sort of thing happens to me every so often (2090A,
>Miniscribe 8051S). I reboot and the machine freezes every so often (usually
>in BindDrivers). If I try to move the mouse, well she don't move at all. If
>I reboot again everything is fine.
>...
>	fred

I've seen this happen on three different A2090's and three different brands of
of SCSI drives.  I would suspect, therefore, that the problem is more likely to
be in the disk driver than in the hardware, (I know it isn't in the disk drives)
since it happens with 2090's and 2090A's.  It only seems to happen around here
right after the system has been powered up, subsequent reboots don't have
any problems. Oh yeah, it doesn't matter how long the power has been on before
booting.  It happened to me last night and I had had the power up for about
five minutes before a stuck in a Workbench disk, so it isn't likely to be a
problem with the disk not being ready.




-- 
=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+
 Jim Levie   REMTECH Inc  Huntsville, Al 
 The opinions expressed above are just that.
 Ph.    (205) 536-8581               email: uunet!ingr!b11!jim

bader+@andrew.cmu.edu (Miles Bader) (01/06/89)

grwalter@watcgl.waterloo.edu (Fred Walter) writes:
> Geez, the same sort of thing happens to me every so often (2090A,
> Miniscribe 8051S). I reboot and the machine freezes every so often (usually
> in BindDrivers). If I try to move the mouse, well she don't move at all. If
> I reboot again everything is fine.

That's exactly what happens wit me... so it seems like it's more
likely the 2090a's fault (or the software or something).

Maybe it's just a race condition somewhere.  [But at least a 2090 is
a lot cheaper than a disk drive...]

-Miles

eric@pyrps5 (Eric Bergan) (01/06/89)

In article <349@b11.INGR.COM.> jim@b11.UUCP (Jim Levie ) writes:
>In article <7546@watcgl.waterloo.edu> grwalter@watcgl.waterloo.edu (Fred Walter) writes:
>>In article <IXjeaqy00UkaMkKllB@andrew.cmu.edu> bader+@andrew.cmu.edu (Miles Bader) writes:
>>>Hi, lately, either my hard disk (q80s) or my controller (2090a) has
>>>been acting very strangely: If I reboot, the reboot succeeds to the
>>>point where the hard disk normally takes over (just after
>>>binddrivers).  But at that point, the drive light will turn on and
>>>stay on, with seemingly no disk activity (that I can hear), and the
>>>system hangs at that point [I've waited up to 15 mins. for something
>>>to happen].
>>
>>Geez, the same sort of thing happens to me every so often (2090A,
>>Miniscribe 8051S). I reboot and the machine freezes every so often (usually
>>in BindDrivers). If I try to move the mouse, well she don't move at all. If
>>I reboot again everything is fine.
>
>I've seen this happen on three different A2090's and three different brands of
>of SCSI drives.  I would suspect, therefore, that the problem is more likely to
>be in the disk driver than in the hardware, (I know it isn't in the disk drives)
>since it happens with 2090's and 2090A's.

	I also was having these kinds of problems, with a 2090 and a ST-251
(note: not SCSI). I actually took the machine in for service, and was told
that it was the hard drive.

	I didn't really want to believe this, but I did put in a different
ST-251, and the problem did go away. (I had previously done various things
such as reseating the cables, etc, so it is unlikely that the problem
was fixed just by unplugging and plugging in boards, cables, etc.)

	I still want to do some more tests to see if the first drive
still fails if both drives are connected.


					eric
					...!pyramid!eric

kim@uts.amdahl.com (Kim DeVaughn) (01/11/89)

In article <IXjeaqy00UkaMkKllB@andrew.cmu.edu>, bader+@andrew.cmu.edu (Miles Bader) writes:
> Hi, lately, either my hard disk (q80s) or my controller (2090a) has
> been acting very strangely: If I reboot, the reboot succeeds to the
> point where the hard disk normally takes over (just after
> binddrivers).  But at that point, the drive light will turn on and
> stay on, with seemingly no disk activity (that I can hear), and the
> system hangs at that point [I've waited up to 15 mins. for something
> to happen].  If I reboot with C-A-A at this point, it boots just fine.
> But if I get a guru, the cycle starts over again....

Three of us here at Amdahl got 2090A/Quantum 80S combos over the Holidays,
and are all experiencing slightly different behavior on COLD starts (i.e.,
power-on boot).

I *always* end up getting the good ol' "Insert WB v1.3" hand holding up a
floppy requester.

Kevin gets behavior similar to yours ... part way thru the s-s (or just a
solid white screen), followed by a hang.

Vern's autoboots all the way up quite nicely (or has thus far ... ).


Quantum specs the 40S/80S drives as having a typical spin-up time of 13 sec,
with a nominal max of 15 sec, and a worst case max of 18 sec.

My *SPECULATION* is that one of these is real close to the timeout that's
built into the hddisk driver code (or is that a 2090A h/w timeout ?), and
that your drive and Kevin's are right on the edge.  My drive is a little
on the slow side WRT spin-up, and Vern's is on the faster side.

I'd really like to know what delay value is used by "hddisk" if that is in
the driver PROM, and *where* it is so I can burn a new PROM with an increased
value.

Also, I'd like to encourage CBM to up the timeout value a bit when/if they
rev this code ... to maybe 30 sec (?).  It *is* rather irritating as is!


Also, as Kevin has mentioned in another article, the LED does not work when
hooked up to either J4 or J5 on the 2090A.  Works fine from the two pins on
the 80S itself, but that won't do me much good when I add a 2nd drive (also
Kevin's and Vern's drive came with a faceplate attached to the drive, and a
LED soldered to the drive's pins).

Any chance this problem can also be addressed with a new rev of the ROM'd
driver code?



Aside from these two problems, I'm VERY pleased with this combination.  It's
*VERY* fast, and so quiet that you seldom know it's there!  Turns Amy into a
whole new machine!!!  Many kudos to whoever it was that did the 2090A board,
the driver code, and Steve for the FFS (and anyone from Quantum whose out 
there).


/kim

-- 
UUCP:  kim@amdahl.amdahl.com
  or:  {sun,decwrl,hplabs,pyramid,uunet,oliveb,ames}!amdahl!kim
DDD:   408-746-8462
USPS:  Amdahl Corp.  M/S 249,  1250 E. Arques Av,  Sunnyvale, CA 94086
BIX:   kdevaughn     GEnie:   K.DEVAUGHN     CIS:   76535,25