My kernel panics suck

[Garrett Cooper]

On Fri, Jul 2, 2010 at 5:40 PM, William D. Colburn (Schlake)
<schlake at gmail.com> wrote:
> On Fri, Jul 2, 2010 at 5:49 PM, Jeremy Chadwick
> <freebsd at jdc.parodius.com> wrote:
>> Ripple or dirty power coming from a UPS -- because not all of them clean
>> things up -- could cause all sorts of chaos hardware-wise too, and in
>> some cases permanent damage.
>
> The UPS is actually my biggest suspect.  It was an expensive one, but
> it is well over five years old now.  As I told someone privately,
> though, I have a broken ankle, and this is my disk server with about
> 9T of spinning disk in it.  It is hard for me to move currently.

If your system // storage controller has an event log (IPMI/RAID?) I
would check to see what's reported by the system at the time of
failure.

Some servers also record events into the BMC (going back to IPMI) and
display trouble codes on the LCDs (like Dell).

It could be an issue with the power supply as well -- see whether or
not you can spin up the disks a little slower (there are knobs in the
kernel you can set to slow down detection so things work), or take
some of the RAID members/volumes offline before booting up so you can
power them down safely?

If your system has a redundant power supply, try using the other one
to see whether or not the issue persists.

As for ruling out the UPS, try plugging the machine straight into the
wall -- does it work?

HTH,
-Garrett