On Jul 15, 2005, at 2:49 AM, Sebastian wrote:
> we have a Cluster with 5 nodes, each node has Dual Opterons on Board.
> Actually we use --with-rpi=sysv as rpi modul, but because we are
> interested
> in using SGE we have to change the rpi modul to --with-rpi=tcp, because
> otherwise the shared memory segments and the Semaphore Arrays don't
> become
> cleared with the "tight integration"-method.
I'm afraid that I know little of the SGE tight integration (other than
knowing that it exists). If you can't use any other RPI other than
tcp, that would be a major limitation of the tight integration.
> Now i want know if the speed between this to modules is different and
> how
> much?
> Does the rpi-modul tcp make much more overhead than the sysv modul?
Robin posted some good links to examine.
But it also depends on your application -- if you do infrequent
communication, or rarely communicate with the processes on the same
node, or perhaps do frequent global synchronizations, then the speedup
from the sysv or usysv RPIs won't help you much. But if you frequently
communicate with your peer processes on the same node and have few
other limiting factors, then the faster speed of the sysv / usysv RPIs
will help you.
--
{+} Jeff Squyres
{+} jsquyres_at_[hidden]
{+} http://www.lam-mpi.org/
|
