[twinspop]

Fookin' 'ell. My XServe has rebooted twice in the last week. Spamprobe was running the frist time. (The second one I failed to catch with syslogd because I was mucking with syslog.config. :-( )

The log clipping is below. This is not good. A userland program running unpriv'd should NOT be able to take the machine down! Any advice? Will support @ apple be useful? (their support hasn't ever impressed me when I've called before. Sort of lowest common denominator. Is it different for Xserves?)


Date/Time: 2003-08-28 20:10:02 -0700
OS Version: 10.2.6 (Build 6L60)
Host: myhost
Command: spamprobe
PID: 23392
Exception: EXC_BAD_ACCESS (0x0001)
Codes: KERN_INVALID_ADDRESS (0x0001) at 0x69a3339b Thread 0
Crashed: #0 0x0034aca4 in __memp_sync_int #1 0x0034ab34 in __memp_fsync #2
0x002fab7c in __db_refresh #3 0x002fa8f8 in __db_close_i #4 0x002fa8a0 in __db_clo
se #5 0x0000845c in FrequencyDBImpl_bdb::close() (FrequencyDBImpl_bdb.cc:131) #6 0x0
0009980 in FrequencyDBImpl_dual::close() (FrequencyDBImpl_dual.cc:53) #7 0x00006440 in
main (spamprobe.cc:62) #8 0x000025a0 in _start (crt.c:267) #9 0x00002420 in start
PPC Thread State:
srr0: 0x0034aca4
srr1: 0x0000d030
vrsave: 0x00000000
xer: 0x00000000
lr: 0x0034ac00
ctr: 0x90004a00
mq: 0x00000000
r0: 0x69676874
r1: 0xbffff590
r2: 0x22000284
r3: 0x0

[twinspop]

Ended up calling the support line. He recommended taking out the extra 512MB module I installed. Said the msg made it sound like bad memory was the cause. So far so good-- no reboots for 3.5 days:

~ (31) % w
8:49AM up 3 days, 15:29, 1 user, load averages: 0.00, 0.00, 0.00

[Buzzfish]

we had a similar problem with ours, rebooting for no reason. but it ended up being a loose processor board, probably shook free during shipping.

Ever since we just reseated that sucker its been fine.


hope that helps,
Dan