[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]
Re: IPX system crash (no log data)
On 10/3/2011 14:04, AGC wrote:
On 9/23/2011 14:43, matthew green wrote:
I'm trying to figure out the cause of a system crash/halt and am having
quite a bit of trouble so I could use your advice.
This is the IPX that is running just ntpd plus the PPS enabled Generic
kernel. Nothing else runs on the system but it seems that after about a
week of running, it just stops. It is unresponsive to network
connections and the console doesn't respond either (sitting at the login
message). I have to Stop-A and boot to restore the system.
None of the normal logs (messages, cron, etc.) and seem to indicate
anything out of the ordinary. No messages are emitted on the console
either. Are there any other logs that NetBSD maintains that might give
me a clue about what happened?
so, when you use stop-A what prompt comes up? db> or ok? from
here, you can get more info (either one) but it depends on which.
if it isn't ddb (db>) then it might be better to build a kernel
with DDB enabled and go from there.
from "db>", you can type "bt" to get a back trace. from "ok" you
should be able to "ctrace" to get the same, but i don't remember
what the IPX firmware supports.
Ok, a heads up: the system crashed sometime last night (once again,
about eight days after rebooting). I'm not in front of the machine at
the moment to provide a trace but one will be coming to the list shortly.
In the mean time, is there anything any of you think I should dump
besides the ctrace? I'll have to transcribe the trace to email by hand
but if there's anything else that will provide more clues, I'll do
Well, it appears that if I wait too long to get to the console I can't
even issue a Stop-A and drop out into the OBP prompt. The system is
just locked hard. Last time it happened I was able to Stop-A into the
OBP and reboot. This time I had to power cycle which obviously kills
any chance of performing a back trace.
Suggestions for instrumenting in case I get locked out again? Otherwise
I'll just have to reboot and try to catch the system again when it
crashes and back trace quickly.
Main Index |
Thread Index |