I'm guessing that the ncpus=1 is what's throwing that off. Remove it and see what happens.
Cheers, -Nick On Fri, Nov 21, 2008 at 7:34 AM, Wu Rongqin <[EMAIL PROTECTED]> wrote: > Dear all, > I have setup a cluster running on Centos5.2. The hardware: Dell poweredge > 1950 for compute node with 64-bit Quad-Core Intel(R) Xeon(R) processors. > That > is, in each node there are 2 physical cpus with 4 Processors in each. > I specify np=8 in server_priv/nodes file for each node. in maui.cfg I just > specify NODECFG[nodex] MAXJOB=8. > The job submission script looks like: > #!/bin/sh > > #PBS -q serial > #PBS -N Serial_Job > #PBS -l nodes=1:ppn=1 > #PBS -l ncpus=1 > #PBS -V > #PBS -o output > #PBS -e error > > ./a.out > > However, when I submit 8 jobs to a node, it seemed that the 8 jobs are > crampped into only one processor, not one job one processor ! > > Anyone can tell me how? > > Regards > > RQ Wu > > _______________________________________________ > mauiusers mailing list > [email protected] > http://www.supercluster.org/mailman/listinfo/mauiusers > >
_______________________________________________ mauiusers mailing list [email protected] http://www.supercluster.org/mailman/listinfo/mauiusers
