[Beowulf] Estimating cluster power consumption

Jim Lux James.P.Lux at jpl.nasa.gov
Thu Dec 15 07:14:04 PST 2005


At 05:28 PM 12/14/2005, r reynoldson wrote:
>Hi all. I'm trying to track down some threads from
>this group that date back to the fall of 2003 (Sept -
>Nov). These don't seem to be in the Beowulf mailing
>list archive. In particular, I read in Jeff Layton's
>"Cluster Environments" article
>(http://www.clustermonkey.net//content/view/40/32/)
>that some cluster power considerations were discussed
>in these threads.
>
>Alternatively, if someone here has some insights and
>can spare a few moments, I have a few questions.
>
>First I should say that I just have a little
>experimental cluster (5 nodes including master) right
>now, but have an opportunity to more than double this
>size very cheaply. The compute nodes are a mixture of
>366 - 450 Mhz boxes, 128 - 256 Mb ram. Currently they
>have hard drives, but I'm seriously considering going
>diskless (for the learning experience, amoung other
>things). Also, the compute nodes are just boxes (no
>monitor, keyboard, or mouse). The master node has an
>1800 AMD cpu, 512 ram, and 60 Gb hard drive. I'll be
>mainly using the cluster for number crunching
>experiments (primality testing, various integer
>factorisation implemetations, etc). A typical cluster
>run could last anywhere from a day to several weeks
>(possibly months?).
>
>My house, which contains my cluster room, is rather
>old. Looking at my breaker panel, it looks like my
>cluster room shares a 15 amp breaker with an adjoining
>room (sad, I know).
>
>Questions
>1) About how many of the above mentioned nodes would I
>be able to safely run?

It's the rare single box that draws more than 200W (unless you've got quad 
xeons or something) without the monitor. You generally don't want to load a 
branch circuit to more than 80% of the rated capacity (0.8*15 = 12A, here) 
12A is about 1200W (actually more like 1400, but fudging on the low side 
here is a "good thing").. so, you can run all 5 boxes on the 15A 
circuit.  (on a more practical note, I used to do this very thing, run 5 
386,486, and Pentium 1 boxes on a single branch circuit in my apartmen, it 
was the LaserJet III that caused the problems, especially if I decided to 
vacuum)

>2) Is there any chance of a fire hazzard if the
>breaker is overloaded or will the breaker just trip?

The breaker "should" just trip on an overcurrent.  However, if the breaker 
is a gazillion years old it might not.  Also, if the wall receptacle is old 
and decrepit (most particularly, the wires on the back of the receptacle), 
then you could be drawing a normal amount of current (i.e. 10A), and the 
receptacle will still get hot enough to cause problems.  Unusual, but it 
happens.

>3) How does one estimate how many nodes they can
>safely run?

Buy yourself a Kill-A-Watt for $30 and actually measure the power consumed. 
(http://www.efi.org/ is one source, there are others)

>4) Would running the cluster diskless help much with
>power consumption? What about no CD and floppy drive?

Every little bit helps

>5) How much juice should be flowing to the room to run
>a cluster of 16 such nodes?

Figure 200W/node (for lack of a better estimate) = 3200W.  3200W/110V = 
29Amps  29A * 1.2 (design capacity factor) = branch circuit capacity of 34 A.

>6) What's the least expensive way of getting more
>electricity to the room?

Extension cords from another branch circuit <grin>.


>I can't do it myself (at
>least I think I can't). Which is the more desirable --
>increasing the existing breaker amperage (if possible)
>or getting a second breaker box (or replacing the
>existing box with a bigger and better box)? The house
>has copper wiring in the basement, which may affect
>the cost (I also live in Canada too, eh!).

A lot depends on what else in the house needs to be fixed when they get in 
to make the changes.  Call your local electrician for an estimate, and they 
can tell you what the tradeoffs are.

>7) What are some good references to answer such
>questions? I'd even spend money on a book, if it were
>recommended highly enough.

Read RGBs online beowulf book at the duke brahma site.  He has a whole 
chapter on infrastructure (power, HVAC, etc) issues.


>Thanks in advance
>Rob

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