On Fri, 27 Feb 2004, Jeff Squyres wrote:
> Unfortunately, we do not plan to support this in the 7.0 code base.
Sure, I was thinking at me doing it :-)
> This is used in all the RPI modules, so LAM would have to be
> recompiled, and the application programs would need to be relinked.
That is not really an option for me now... I can give up feedback on
7.1-cvs but I can't put into production. So then I have to go back to
fighting with the queueing system to give me more flexibility. I don't
mind for the moment having both lamd and computation communication on
the same network, but adding the queueing system will be overkill,
while leaving the other network with almost no traffic :-(
Many thanks for the details about hostmap implementation. Maybe
they'll be useful to somebody else on this list...
> It's roughly equivalent, either way
Well, I prefer to do things with pipes rather than writting temporary
files that will only be read immediately afterwards and only once and
then have to be deleted. With files, you always have to pay attention
to how and where you create them; with pipes, the OS takes care of
everything. :-)
--
Bogdan Costescu
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_[hidden]
|
