What is a Beouwlf? (was HD cloning)
Walter B. Ligon III
walt at parl.ces.clemson.edu
Wed Dec 6 07:41:24 PST 2000
> On Wed, 6 Dec 2000, Bruce Janson wrote:
> > And to you Robert Brown: speak for yourself please when you say
> > (in your message of Sun, 3 Dec 2000 16:03:39 -0500 (EST)):
> > A beowulf is a high performance computing
> > cluster, not a data or web server cluster.
> > This kind of supercomputer elitism, this fascination with fine-
> > grained parallelism and linear speed-up has held back the progress
> > of single system image multicomputing for long enough. I disagree
> > with your claim, so much so that I wouldn't even fight for your
> > right to make it (well, not with much conviction).
> Whew! Harsh words! I don't know how you read "fine-grained
> parallelism" into the phrase "high performance computing cluster" -- I'm
> actually an embarrassingly parallel Monte Carlo kind of guy (and linear
> speedup is what you get when you're NOT fine-grained, so I make out just
> fine there;-). I'd therefore be deeply hurt if I weren't wearing
> teflon-coated asbestos over a kevlar vest;-)
[ many eloquent and to the point words removed to save space ]
> Is this fair? Am I being crazy here? I'm not even a Real Computer
> Scientist, so casting me as a fine-grained computing fiend who is
> somehow obstructing "true, unbridled" beowulf development is just not
Robert, you are not crazy, and you have been a part of this community long
enough to deserve quite a bit more respect than you were offered here. List
readers in general (not just this list) are well advised to lurk long enough
to know something about the community they are conversing with before
launching unprovoked flames. Personally I like that this list normally
doesn't have such flames. Maybe Mr. Janson should consider this before his
next post of this ilk.
A Beowulf is a high performance parallel computer. Period. This definition
has been standing for a good five years or more, there is no good reason
to change it now. Not all clusters are Beowulfs.
There are other OVERLAPPING communities who are interested in clusters
for other purposes. This is a good thing. These communities follow
the Beowulf group because many Beowulf technologies are also useful for
these other purposes. Some others are not. This is also a good thing.
This does not mean that everything is a Beowulf and a Beowulf is everything.
This would not be a good thing. In particular check out the "Extreme Linux"
That said, this list has always proved flexible in the kind of stuff posted
here. This is a good thing. Posts such as the one quoted above only hurt
that relationship, and thus should be avoided. They are not good things.
Now to a couple of technical points (if anyone is still reading): Beowulf
computers are no good for fine-grain parallelism (according to what most
people I know think of as "fine grain"). Bewulfs are coarse grain, or
possibly medium grain if you invest in some expensive networking hardware.
Next, single system image multicomputing has been PUSHED by the high
performance computing community. Case in point is the new Scyld Beowulf
Robert, consider yourself vindicated!
Dr. Walter B. Ligon III
More information about the Beowulf