Subject: random reboots with raid drive access
To: None <current-users@netbsd.org>
From: Joel Leyh <hashbang@gmail.com>
List: current-users
Date: 04/19/2005 00:30:09
I have Netbsd 2.0 on a Dell Poweredge 4300 (single proc), 512 ram and:

aac0 at pci2 dev 10 function 0: Dell PERC 2/QC
aac0: interrupting at irq 5
aac0: StrongARM SA110 at 233MHz, 128MB cache, required battery
present, kernel 2.8-0
ld0 at aac0 unit 0: RAID 0 (Stripe)
ld0: 4065 MB, 2065 cyl, 64 head, 63 sec, 512 bytes/sect x 8326144 sectors
ld1 at aac0 unit 1: RAID 0 (Stripe)
ld1: 17354 MB, 8815 cyl, 64 head, 63 sec, 512 bytes/sect x 35542272 sectors

Sometimes when I do a rm -rf pkgsrc or a tar zxf pkgsrc.tar.gz or a
make clean on a large package like gcc or postgres, my machine will
reboot (doesn't matter which drive.) I do not see anything flash on
the screen, nor can I find any mention of a core dump or crash once
the machine is back up. Every single time after this "crash", I get a
large amount of UNKNOWN FILE TYPE, UNALLOCATED, and corrupt
directories and files when I run fsck_ffs. This is annoying, since I
will randomly loose mkdir, or /etc/syslog.conf, etc. Is this an issue
with the raid controller? This has happened both with the generic
kernel and my own custom build.

regards, Joel