[Beowulf] Emergency Power Off

Jim Lux James.P.Lux at jpl.nasa.gov
Sun Mar 18 12:16:00 PDT 2007

At 10:36 AM 3/18/2007, Robert G. Brown wrote:
>On Sun, 18 Mar 2007, Jim Lux wrote:
>>There's a huge "small office and retail" sort of market for 
>>middling sized UPSes, and APC and the like would like to sell into 
>>that market as a "plug and play" product.  Also, this is where the 
>>local AHJ can play a role.  THEY can say, we don't care if your 
>>installation isn't in a room that meets ALL of the requirements of 
>>645, we want you to provide a disconnecting means anyway.
>I could wish that they made this a whole lot simpler.  And perhaps
>standardized so it would be cheap as well as easy.

The whole "multiple sources of power" thing is becoming a more 
noticeable issue.  Consider, for instance, a building with an array 
of solar panels.. electrically live all the time! The code has had a 
series of revisions over the past few years to keep up with this.

Standby and portable generators are another area of some changes in 
the recent codes, especially with respect to grounding and "neutral" 
and what kinds of transfer switches.

There's a sort of tension between the factions that come up with 
unlikely scenarios and the factions that don't want to require 
excessive (in a cost/benefit sense) and expensive remedies for 
unlikely events, and through such tension we get to the code we've got.

Usually, you need some particularly egregious event to catalyze a 
substantial change (The MGM Grand fire was one such event.  The very 
wide use of portable generators after hurricanes was another.)

There's also a (very wise in most cases) tendency to not impose code 
requirements on situations where everything is under the control of a 
single person, including both execution and consequences.. i.e. if 
you want to electrocute yourself on your 10 acre farm, that's your 
own business.  But, if you have 50 employees, you'd better make sure 
they're safe.

I would like to see a more consistent and integrated way to handle 
the sorts of issues you bring up with UPSes and the like.  But to a 
certain extent, we're faced with the classic beowulf dilemma 
(clusters, not nordic heros)... you want to leverage off consumer 
gear, which by and large is not intended to operate in an industrial 
setting (either in terms of environment, maintainability and 
reliability, FCC Part 15, or anything else).  What's perfectly ok for 
a consumer to have sitting in their living room in qty 1 may not be 
ok in a wiring closet with qty 50.

Hmm... you know, of course, that plugging one plug strip into another 
one is strictly forbidden?  And there's a reason that the cords on 
most plug strips are only 6 ft long?  That's driven by the "expected 
usage" which in turn defines the "credible scenarios" which in turn 
defines the "required risk mitigations" which in turn defines "who 
loses in a lawsuit if there's a fire"

Of course, it's a bit unrealistic to expect the guy or gal in the 
orange or blue vest at the big-box store to know all these things and 
advise the clamoring hordes of customers accordingly.

And, that white paper from APC (which, by the way, has a revised 
version for 2005 that talks more about the risks of deliberate EPOing 
as sabotage) is pretty short on practical details, while it's great 
at showing drawings of buttons.

>I'm going to try to figure out how to put an EPO in this room for less
>than a fortune, even though I'm pretty sure that it isn't strictly
>required.  Even if it is just one that controls the UPS's -- so that
>the building power can be cut externally and one can bop off the UPS's
>on entering the room.
>The building power is tricky enough as it is -- they have a gas
>generator that kicks in transparently on loss of grid power (it's a
>medical clinic).  That is not my thing, though.
>    rgb
>Robert G. Brown                        http://www.phy.duke.edu/~rgb/
>Duke University Dept. of Physics, Box 90305
>Durham, N.C. 27708-0305
>Phone: 1-919-660-2567  Fax: 919-660-2525     email:rgb at phy.duke.edu

James Lux, P.E.
Spacecraft Radio Frequency Subsystems Group
Flight Communications Systems Section
Jet Propulsion Laboratory, Mail Stop 161-213
4800 Oak Grove Drive
Pasadena CA 91109
tel: (818)354-2075
fax: (818)393-6875 

More information about the Beowulf mailing list