The FNN (Flat Neighborhood Network) paradox
Georgia Southern Beowulf Cluster Project
gscluster at hotmail.com
Fri Feb 23 14:15:28 PST 2001
This is the primary drawback I can think of immediately, however, as you've
pointed out the method I described is "possible" though it may not be
significantly useful. I just wish to know if it is really, really possible
and if so can it be used to study the software and situation to develop a
much better solution (faster loading module or a daemon that can monitor for
this situation). I'm by far not experienced in these matters, but I think
the thought has merit. Happy Hacking to whoever tries this (would, but no
FNN/AFN cluster to do it with).
>From: Bogdan Costescu <bogdan.costescu at iwr.uni-heidelberg.de>
>To: Georgia Southern Beowulf Cluster Project <gscluster at hotmail.com>
>CC: Beowulf List <beowulf at beowulf.org>
>Subject: Re: The FNN (Flat Neighborhood Network) paradox
>Date: Fri, 23 Feb 2001 13:50:30 +0100 (CET)
>On Thu, 22 Feb 2001, Georgia Southern Beowulf Cluster Project wrote:
> > 3. M2 loads the channel-bonding module (all NICS are on different
> > and sends a reply over the AFN to M1 that it is ready for transmission.
>Loading a module is not an instant operation. You can send lots of normal
>(i.e. on a single link) traffic before the module becomes operational. If
>your software is sensitive to latency (apart from being sensitive to
>bandwidth which is the reason why you would use bonding), you loose.
>But if you're doing something coarse-grained (like image rendering), it
>might fit well.
>IWR - Interdisziplinaeres Zentrum fuer Wissenschaftliches Rechnen
>Universitaet Heidelberg, INF 368, D-69120 Heidelberg, GERMANY
>Telephone: +49 6221 54 8869, Telefax: +49 6221 54 8868
>E-mail: Bogdan.Costescu at IWR.Uni-Heidelberg.De
Get your FREE download of MSN Explorer at http://explorer.msn.com
More information about the Beowulf