[Beowulf] Carpet glue and live clusters?

[Robert G. Brown]

On Mon, 29 Aug 2005, David Mathog wrote:

> The classroom next to my current machine room is going to be
> carpeted soon via the laying down of adhesive followed by the
> placement of carpet squares.  Unfortunately this classroom shares the
> same A/C as the machine room so the fumes will circulate through
> the computers located there.  And no place else, it's just these
> two rooms on that A/C unit.
> 
> I'm thinking  when the day comes I'll shut down the computers
> and the A/C as well.  Then the only fumes that would reach
> the machine room would be those that diffused through the
> inactive ducts between the two rooms.  When that low level
> of fumes enters the computers they will find only cold
> surfaces, and so who-knows-what chemistry that might take
> place on surfaces with temps >60C will be avoided.
> Also with the A/C and computers off there's no risk of a spark.
> 
> Am I being too conservative?  Seems to me like a case of better
> safe than sorry but I've heard one local opinion that there's
> nothing to worry about.

I think that you've got nothing to worry about.  Remember, the computers
themselves are not cold surfaces, they are in fact rather warm.  The
"fumes" are likely to be water (probably mostly water) and some
volatiles.  The volatiles will be in very small concentrations and to
the extent that they DO precipitate out at chller temperatures (which
might well be "not at all" except for the water) they'll precipitate out
in the chiller and be drained away, or maybe sit on the walls for a day
or two as a monomolecular layer until air ciculation carries them away.

Even the amount that precipitates out in a warm, wet, chemically active
environment such as your lungs aren't terribly likely to cause immediate
problems, especially if the overall air circulation is decent (so new
air is constantly being mixed in to re-evaporate and carry off the
volatiles).

I'd worry more about the dust that would accompany this process.  Carpet
(new or old) is plumb full of tiny fiber fragments that are easily
carried about in air and which will actually precipitate out in a layer
a mm thick on nearby surfaces when they lay the carpet.  If they have to
do any surface prep (roughening the surface with a sander or the like)
before laying it, that will exacerbate the problem.  Pulling up the old
carpet produces fiber fragments and dust mites and more.  Computers DO
hate dust-bunny dirt as it clogs the fans and coats the heatsinks with
an insulating layer (both bad things).

So I think you'd be fine if you just make sure you've got good,
functional filters in your air delivery system so that any air taken up
in the carpeted room is filtered before delivery (mostly to get the dust
out, as the volatiles will mostly go right through or stick to the
filters or chiller tubing).  And then change the filters right
afterwards, if a visual inspection warrants it.

   rgb

> 
> Thanks,
> 
> David Mathog
> mathog at caltech.edu
> Manager, Sequence Analysis Facility, Biology Division, Caltech
> _______________________________________________
> Beowulf mailing list, Beowulf at beowulf.org
> To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf
> 

-- 
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