You used it correctly. Remember, all that cpu number is telling you is the percentage of use by that process. So bottom line is: we are releasing it as much as we possibly can, but no other process wants to use the cpu, so we go ahead and use it.

If any other process wanted it, then the percentage would drop and the other proc would take some.

On Dec 4, 2009, at 8:13 AM, Nicolas Bock wrote:

On Fri, Dec 4, 2009 at 08:03, Ralph Castain <rhc@open-mpi.org> wrote:


It is polling at the barrier. This is done aggressively by default for performance. You can tell it to be less aggressive if you want via the yield_when_idle mca param.


How do I use this parameter correctly? I tried

/usr/local/openmpi-1.3.4-gcc-4.4.2/bin/mpirun -np 3 --mca mpi_yield_when_idle 1 ./master

but still get

nbock    20794  0.0  0.0  53980  2344 pts/2    S+   08:11   0:00 /usr/local/openmpi-1.3.4-gcc-4.4.2/bin/mpirun -np 3 --mca mpi_yield_when_idle 1 ./master
nbock    20795 96.7  0.0 159096  3896 pts/2    R+   08:11   1:10 ./master
nbock    20796  0.0  0.0 158960  3804 pts/2    S+   08:11   0:00 ./master
nbock    20797  0.0  0.0 158960  3804 pts/2    S+   08:11   0:00 ./master
nbock    20813 88.1  0.0 158964  3892 pts/2    R+   08:12   0:08 ./slave arg1 arg2
nbock    20814 86.9  0.0 158964  3908 pts/2    R+   08:12   0:08 ./slave arg1 arg2
nbock    20815 87.5  0.0 158964  3900 pts/2    R+   08:12   0:08 ./slave arg1 arg2

nick

_______________________________________________
users mailing list
users@open-mpi.org
http://www.open-mpi.org/mailman/listinfo.cgi/users