Dear all,
is "-np N" parameter needed for mpirun when running jobs under SGE
environment? All examples in
http://www.open-mpi.org/faq/?category=running#run-n1ge-or-sge
show that "-np N" is used, but in my opinion it should be redundant:
mpirun should determine all parameters from SGE environment anyway. If
there is "no need to specify what nodes to run on" (as stated in FAQ),
it should not be necessary to specify the number of processors either.
I ran some tests and it seems that mpirun correctly gets number of
processors from SGE, but I would like to have some more solid
confirmation.
You might want to use a smaller number of processors than those made
available by SGE.
Thanks for replying. I can imagine that in some special cases it might
be useful to request N processors from SGE and than use M<N; in such
cases I would specify "-np M" for mpirun, of course.
But this does not answer the question - if I just want to use as many
processors as I request from SGE - do I need to specify "-np N" for
mpirun or not?
Jiri Polach
Ricardo Reis
'Non Serviam'
PhD/MSc Mechanical Engineering | Lic. Aerospace Engineering
Computational Fluid Dynamics, High Performance Computing, Turbulence
http://www.lasef.ist.utl.pt
Cultural Instigator @ Rádio Zero
http://www.radiozero.pt
http://www.flickr.com/photos/rreis/
contacts: gtalk: kyriu...@gmail.com skype: kyriusan
Institutional Address:
Ricardo J.N. dos Reis
IDMEC, Instituto Superior Técnico, Technical University of Lisbon
Av. Rovisco Pais
1049-001 Lisboa
Portugal
- email sent with alpine 2.00 -
_______________________________________________
users mailing list
us...@open-mpi.org
http://www.open-mpi.org/mailman/listinfo.cgi/users
