[CentOS] Power failure diagnosis?

Wed Aug 18 18:57:34 UTC 2010
Mark <mhullrich at gmail.com>

This morning between 07:10 or so and 08:10 I discovered that my PC had
shut itself off.  I thought it was due to a power failure becuase I
never turn it off without a good reason, and it had been running (and
playing sounds) until it wasn't.  My roommate/landlord tells me that
the only anomaly he observed was that our router had lost its internet
connection (via a cable modem, which was still running fine), and his
computer, a laptop, and clocks all seemed to be running normally.

I checked all the logs in /var/log and there's nothing in there other
than the restart - no temp warnings or anything.  The machine has been
up and running smoothly since the last time I shut it off (to move), a
week ago last Friday  (8/6).

I don't have a UPS or BB, so is there any way that I can diagnose this further?

Here's another weird thing - 'dmesg | grep Aug' shows this:

Linux version 2.6.18-194.11.1.el5 (mockbuild at builder10.centos.org)
(gcc version 4.1.2 20080704 (Red Hat 4.1.2-48)) #1 SMP Tue Aug 10
19:05:06 EDT 2010
EDAC MC: Ver: 2.0.1 Aug 10 2010
EDAC amd64_edac:  Ver: 3.2.0 Aug 10 2010

That's probably when I last restarted after installing the new kernel.  But:

$ date
Wed Aug 18 11:52:23 PDT 2010

And /var/log/messages shows this:

:
Aug 18 06:35:57 marichter dhclient: DHCPREQUEST on eth0 to 192.168.0.2 port 67
Aug 18 06:35:57 marichter dhclient: DHCPACK from 192.168.0.2
Aug 18 06:35:57 marichter dhclient: bound to 192.168.0.100 -- renewal in 5154 se
conds.
Aug 18 06:44:28 marichter smartd[3829]: Device: /dev/sdb, 4294967295 Currently u
nreadable (pending) sectors
Aug 18 06:44:28 marichter smartd[3829]: Device: /dev/sdb, 4294967295 Offline unc
orrectable sectors
Aug 18 08:10:45 marichter syslogd 1.4.1: restart.
Aug 18 08:10:45 marichter kernel: klogd 1.4.1, log source = /proc/kmsg started.
Aug 18 08:10:45 marichter kernel: Linux version 2.6.18-194.11.1.el5 (mockbuild at b
uilder10.centos.org) (gcc version 4.1.2 20080704 (Red Hat 4.1.2-48)) #1 SMP Tue
Aug 10 19:05:06 EDT 2010
:

What's up with that??? (Not the disk issue - that's a Seagate problem
that, fortunately, means nothing, and yes, I've gone through that one
before....)  Why doesn't dmesg show the most recent boot?

marichter 2.6.18-194.11.1.el5 #1 SMP Tue Aug 10 19:05:06 EDT 2010
x86_64 GNU/Linux
centos-release-5-5.el5.centos.x86_64

Thanks.

Mark