[Nut-upsuser] staging shutdowns

[Darin Perusich]

Hi Arnaud,

On Fri, Mar 9, 2012 at 6:00 PM, Arnaud Quette <aquette.dev at gmail.com> wrote:
>
> I'm very interested in helping you there, and complete documentation
> (Typical setup...)

Thanks, but I've actually just finished rolling this out to most of my systems.

> To provide you with a suitable, I'll need to know a bit more about
> your setup, Ie number of UPS / PDU, power topology (ie 1 big UPS only,
> redundancy on some groups, brand of HW (for features), ...). The more
> details, the better.

APC Symmetra 40K w/ a network module, snmp enabled.

> if you want to use NUT to accomplish the task, you will indeed have to
> use a combination of:
> - upsmon (1 master per group + 1 global master ; then all other
> systems in groups are slaves),
> - upssched,
> - clone or clone-outlet or dummy-ups,
> - some scripting (shell, python, perl).

There are a mixture of physical and virtual hosts which adds a fair
amount of complexity to the setup. My NUT master polls the APC, and
all the slaves poll it. I'm using upssched and upssched-cmd based on
the examples from http://rogerprice.org/NUT.html to initiate a timed
shutdown of the hosts. I've deployed NUT to the systems using a Puppet
module/class I wrote which allows me to specify when to initiate
shutdown based on whatever criteria I choose for said systems...i.e
shutdown "dev" after 5 minutes, "production" after 10, once all VM's
are down take down vmware, etc.

> I'm thinking for some time now about a "nutmon" that would still be
> simple by default, but that would allow to address everything possible
> in NUT.
> so your kind of feedback would be very useful.
>
> cheers,
> Arnaud
> --
> Linux / Unix Expert R&D - Eaton - http://powerquality.eaton.com
> Network UPS Tools (NUT) Project Leader - http://www.networkupstools.org/
> Debian Developer - http://www.debian.org
> Free Software Developer - http://arnaud.quette.free.fr/