[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: DATACENTER: inductive power measurement
What is available is a special inductive sensor that attaches to a
standard DVM and measures actual current flow. It is non-intrusive. It
is also not cheap (It''l cost more than an esspresso lunch<grin>).
However, measuring actual power usage is not always a good idea.
Especially if you have empty space in the rack. The empty space tends to
get filled with more heat generating equipment. What I tend to do is use
equipment ratings from the manufacturer. If they say that their
equipment uses 3A at 110VAC then I use that to calculate heat
dissipation directly. By definition, when equipment is rated to use 10
watts then that is exactly how much heat it will generate. The
manufacturer rating is a maximum, meaning higher than expected average.
Your actual measurments will always be lower then that number. If they
are not, then you have a problem somewhere (failed component).
However, when your temprature sensors start climbing through the roof,
in your rack, what is happening is ACCUMULATED heat over time. Your rack
isn't venting the heat fast enough. Either reduce the number of
compenents in the rack, or fix the air-flow problem. Sometimes just
re-organizing the equipment, within the rack, to separate adjacent
hot-boxes, may solve your problem.
I don't have time to look up the references just now, but air absorbs
heat at a specific and constant rate. That can be used to calculate the
amount of air volume that would have to be replaced, in a rack, at a
given power usage level. This is the reason that fans are rated in cubic
feet per minute (CFM).
> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]]On Behalf Of Peter Polasek
> Sent: Thursday, June 17, 1999 11:21 AM
> To: [email protected]
> Cc: Peter Polasek
> Subject: DATACENTER: inductive power measurement
>
>
>
> We have seen a significant discrepancy between the actual and
> computed power consumption in a rack (using the the manufacturer's
> average consumption rating for the computation). Therefore, in
> our power analysis, we would like to measure the actual power load
> in addition to computing the theoretical peak and average rates.
> Is there a inductive device that allows the power or current draw
> to be non-intrusively measured (i.e. does not require the equipment
> to be powered down or the breaker to be physically dismantled).
> The 'Fluke Model 33' was recommended to me; however, I'm not sure
> whether that is (or any device) has an inductive pickup that can
> be used for non-intrusive power measurement.
>
> Thanks in advance,
>
> Peter Polasek
> [email protected]
>