I have a server in my environment that has the tendency
to reboot without warning. I checked the logs and the
see the only error listed is the unexpected shutdown
event. I see no other corresponding events!! I have run
a monitoring utility on the box a found that before the
reboot there seems to be a high disk fragmentation %.
Has anyone seen something like this?? I'm not sure if
the drives are going bad or this is just a result of
another problem.

Get the disk diagnostic at the disk manufacturer's website. It'll tell
you if the drive's failing. Also run chkdisk (no parameters) to see what
it reports, and run it with the /f parameter if it reports anything
wrong (but only if the diagnostic reports nothing wrong or fixes
problems it finds.) Also get a good defragmenter like PerfectDisk and
run it (same caveat as with chkdisk, altho PD I believe never puts the
files and index-type structures it's moving around or modifying in a
state where a sudden system stop is dangerous.)

Do these things in that order, except that you might try chkdisk (no
params) first. If the system crashes/reboots during a defrag or chkdisk
with the /f parameter the disk file structures can be badly damaged.

You'd also be wise to get all the data off that disk ASAP, if possible.

And in recovery options remove the check mark next to immediate reboot.
That way at least you'll get a stable BSOD in most cases and there may
be a useful clue displayed.

There are many other possible explanations for the reboots, including
failing RAM, CPU overheating, driver corruption and the like. You should
take a look at Device Manager to see if there's anything there.

Vincent Cruz wrote: