Nut upsuser - Jun 2016 - unexpected shutdown, had to reinstall xorg.server

Greetings all;

I expect that I may not have the shutdown properly configured, but my 
site was subjected to some bad weather related intermittent power losses 
a week+ back, and I have an automatic standby plant, a 20kw, that usuall 
fires up and has the lights back on in 7 or 8 seconds.  So while the 
power folks were reclosing substation breakers to see where the next 
tree limb was shorting things out, the standby was shut down and then 
restarted quickly 3 or 4 times. Somewhere in all the bounces, the UPS, a 
belkin, shut off its AC input without making any advise me noise in the 
logs, which I was tailing at that point.

The net result being that I've had commercial power for 10 minutes or so 
and I am replying to an email message when click and the whole system is 
dead from a low battery shutdown.

So why did the ups do a powerdown?

Restarting the UPS with a long push on its power button was done, and I 
expected a normal reboot to bring things back to life.  Wrong...  No X, 
and virtual crickets in the logs. Actually no xorg.0.log.  /usr/bin/X 
nowhere to be found.  Pointed shotgun at

 apt-get install --reinstall xorg./server*

and it reinstalled 34 packages.  sudo reboot and it was back to the usual 
everything starts on the same screen & has to be moved to put it on the 
screens I am used to mess.

This is a debian wheezy based install, and the system drive is ext4, and 
the 2nd drive is ext3, for amanda's use as backup medium using virtual 
tapes. I should probably make that one an ext4 drive also. Just never 
found my round tuit.

A upsd -V says 2.6.4 which is pretty ancient by now.

Whats the best way to proceed to avoid this in the future?

As an aside, neither of the /etc/nut/*.html files will display the ups 
status, displaying the source text in place of the values expected.

Configuration error?  Main man page for describing that plz.

However a "upsc myups" shows the expected results.

I've asked several questions.  Hopefully they all have an answer.

Thanks all.
 
Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Genes Web page <http://geneslinuxbox.net:6309/gene>

On 06/27/2016 08:08 AM, Gene Heskett wrote:\
> The net result being that I've had commercial power for 10 minutes or
so
> and I am replying to an email message when click and the whole system is 
> dead from a low battery shutdown.
Since that was apparently not initiated by nut, the problem is with the
UPS.  Yes, we've junked several UPSes that would shut off in the middle
of a work day with no commercial power problems at all.

While I would recommend your new UPS be a little better quality than
Belkin (I suggest CyberPower as still cheap, very nut compatible, and
very reliable in our decades of experience), any UPS can fail.   The
best way to protect against that is with dual power supplies - although
that may be overkill for your application, it is pretty much essential
for a server.

On Jun 27, 2016, at 8:08 AM, Gene Heskett wrote:
> 
> As an aside, neither of the /etc/nut/*.html files will display the ups 
> status, displaying the source text in place of the values expected.
> 
> Configuration error?  Main man page for describing that plz.
/usr/share/doc/nut-cgi/README.Debian has step-by-step instructions, although it
is a little light in the web server configuration section (naturally, given how
many are out there). I think the cgi-bin pathname works with the default
configuration for Apache 2.

-- 
Charles Lepple
clepple at gmail