box (a):
dual 330MHz PII, 440LX, 128Mb
Adaptec AHA-294X UW SCSI w/ 4Gb DCAS-34330W
Intel EtherExpress Pro 10/100 @ 100baseT/HD
box (b):
dual 400MHz PII, 440BX, 256Mb
Adaptec AIC-7895 UW SCSI (dual channel w/ 2x 9Gb WDE9100AV)
Intel EtherExpress Pro 10/100 @ 100baseT/FD
DEFPA FDDI @ 100Mbits/s
Both have suffered *hard* hangs with SMP kernels from 2.0.35 (RH5.1)
and 2.0.36pre14. The symptoms are identical: blank screen, no keyboard
LED response, no warning and nothing in the logs. Only seen it happen
with load => 2.
Box (b) is a server feeding 6 Xterminals, running between 100-250 procs
at any one time. Box (a) is a lab seat with ~60 procs. The server
lasts an average of 4 days with a 2.0.35 kernel and I just had my first
hang with 2.0.36pre14 after 11 days of uptime.
After the SMP fix in 2.0.36pre14, I thought I was home free, but there
still seems to be a problem. The lab seat has also suffered the same
symptoms but with longer uptimes of a month or so with 2.0.35.
I've not seen anything similar on any of my UP boxes or even on these
SMP boxes with UP kernels.
I guess I'm not the only one seeing this, having read other mails
stating identical symptoms. Is anyone still looking at SMP deadlocks in
2.0.x? Does anyone have any suggestions?
Cheers,
Wilf.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/