3 May
2001
3 May
'01
8:59 a.m.
All,
Having some problems with my beloved NeXT. I'm *hoping*, perhaps
unreasonably, that they are related to the internal hard drive. I want to
try replacing it with a different one. Here's what I'm looking for:
Very Reliable
Low Power consumption/heat dissipation
3.5 inch
50-pin SCSI
550 Meg < capacity < 2 Gig
Inexpensive
Recommendations? Both drive model, and where to get one? I have an
RZ26 which would probably be perfect, but I hate to use up a DEC drive in a
non-DEC machine, having heard of the trouble folks have had trying to go
the other way. (Still awaiting CD and User ID number from Mentec and
Encompass, respectively.)
Other (possibly pertient) info:
Machine is a 68040 cube, non-turbo. Its drive bays are big enough
to accomodate 5.25 inch drives, but the lower bay contains an optical and
the upper bay contains a floppy *and* the internal hard drive. There is an
external case with a pair more hard drives and some pretty questionable
cabling, but the fault has appeared with that whole setup disconnected
(though not lately).
Internal hard drive (and floppy) are cantilevered between a pair of
plastic plates that are screwed to the inside of the drive bay. The
cantilever is formed by a set of mounting screws that go from the plastic
to the drive screw holes via plastic stand-offs. I'd say it's pretty likely
the drive is electrically isolated from the chassis except for the power
plug ground. I'd say it's extremely likely the drive is thermally isolated
from the chassis, so it's entirely air-cooled. The
fan is circulating air
in the volume where the drive is hanging.
Fault appears during heavy disk activity only. Can happen as a
consequnce of
find / -name .nfs ....
for example, which appears in the daily cron job (sigh). Can also happen when
find / -name .nfs
is entered from a terminal session. Things like
cd /
du
can produce "segmentation fault" errors as well.
/private/adm/messages log file shows unsettling things like:
May 3 10:16:05 mtapley2 mach: unexpected kernel page fault failure
May 3 10:16:05 mtapley2 mach: trap: type 0x410 fcode 5 rw 1 faultaddr
0xb570101a
May 3 10:16:05 mtapley2 mach: trap: pc 0x4001ada sp 0x3fffb0c sr 0x2000
May 3 10:16:05 mtapley2 mach: trap: cpu 0 th 0x101b2030 proc 0x101b1af0
pid 210 pcb 0x101b2230
May 3 10:16:05 mtapley2 mach: traceback: fp 0x111d5fb0
May 3 10:16:05 mtapley2 mach: called from pc 0x040f15c6 fp 0x00000000
4-args 00000001 b570101a 00000000 00000000
May 3 10:16:05 mtapley2 mach: last fp 0x0
May 3 10:16:05 mtapley2 mach: panic: (Cpu 0) MMU invalid descriptor during
table walk
(Is this consistent with a soft hard-drive failure, or am I barking up the
wrong tree entirely?)
Sometimes during attempts to reboot after fault, monitor shows file systems
clean but then I later get a panic - requiring fsck for the next reboot.
Fsck often has to be run repeatedly. (Fsck represents fairly heavy disk
activity. Sigh.)
Fault has not appeared within 30 minutes or so of powering the
machine down and leaving it to cool for 10 minutes or so, so may be at
least partially due to thermal problems.
Advice or recommendations for things to try are welcome. Thanks in
advance for any form of help.
- Mark