[Pw_forum] Code starting trouble on the PC Cluster
akohlmey at cmm.upenn.edu
Tue Aug 22 17:33:53 CEST 2006
On 8/22/06, Shujun Hu <hushujun at mail.sdu.edu.cn> wrote:
> Dear Drs,
> When starting the parallel code pw.x on the cluster (Intel Pentium-D and MPICH2),
> usually i need run the command several times. Nothing appear on the screen for the
> early commands but the last one seems good. I have checked the process queue and
> found that lots of process named pw.x under python were sleeping. I donnt know
without any detail information, it is almost impossible to give
specific advice. you have to help us to help you. none of us
is a psychic and can read your mind (or the one of your computer).
so you should:
a) try out and test the test programs, that come with the
MPICH2 software. if you cannot make them run, you
have to fix your setup, since you won't be able to run
_any_ parallel MPI program.
b) let is know the command line you were using. please check
the manual about running parallel jobs. many parallel environments
do not support reading input files from stdin, so you may need to
use the -input flag. the fact that you have many hung processes
is an indication that something like this might be happening.
hope that helps,
> Shujun Hu
> Pw_forum mailing list
> Pw_forum at pwscf.org
Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu
Center for Molecular Modeling -- University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425
If you make something idiot-proof, the universe creates a better idiot.
More information about the users