Nut upsuser - Sep 2015 - monitoring Eaton parallel/redundant UPSs

I have been happily using NUT to monitor a couple of Eaton UPS. These 
are independent UPSs, each driving one "leg" of the power in our data 
center. In this scenario, I've told my clients they need to have one UPS 
working and to shutdown when the last UPS reaches critical battery.

I have recently learned that our power arrangement will be changing to a 
parallel-fed/redundant configuration. In this arrangement, I think we 
will have two UPSs whose output is routed through a "Bypass Paralleling 
Switchboard". The output of this will feed both legs of the power in our 
data center. In theory, either UPS should be able to carry the load. If 
this is the case, does my monitoring arrangement need to change?

My initial though was "no". If UPS-A and B are both on-line (OL),
I'm
golden. If A is shutdown and B is OL, I'm still above my minimums. If B 
then goes critical, I can perform my server shutdowns. In this case, it 
all sounds fine.

But what about that "Bypass Paralleling Switchboard"? I suspect it is 
possible that it could be used to ignore the input from UPS-A and rely 
only on UPS-B to handle the load. If I'm only monitoring the status of 
the UPSs, how will I know that having UPS-A OL and 100% is irrelevant? 
Does anyone know how to incorporate the state of this "paralleling" 
component into my NUT monitoring?
-- 
    Do things because you should, not just because you can.

John Thurston    907-465-8591
John.Thurston at alaska.gov
Enterprise Technology Services
Department of Administration
State of Alaska

On Sep 2, 2015, at 5:54 PM, John Thurston <john.thurston at alaska.gov>
wrote:
> 
> But what about that "Bypass Paralleling Switchboard"? I suspect
it is possible that it could be used to ignore the input from UPS-A and rely
only on UPS-B to handle the load. If I'm only monitoring the status of the
UPSs, how will I know that having UPS-A OL and 100% is irrelevant? Does anyone
know how to incorporate the state of this "paralleling" component into
my NUT monitoring?
Hi John,

I am not qualified to offer any advice about that bypass system (I try not to
touch anything bigger than 120V/15A), but since nobody else has piped up yet: if
you can get details from your clients about how that works conceptually, we can
help map that into a NUT configuration.

The big question in my mind is whether there is some way to programmatically
determine what state the circuit is in. If not, there is still probably a way to
manually use the dummy-ups driver to add in a phantom UPS representing the
parallel-fed state.

-- 
Charles Lepple
clepple at gmail