Hi All.  Thanks in advance.

I'm trying to send many, relatively small jobs to a 10 node cluster with
4 CPUs per node.  The code that I have uses pthreads and runs nicely on
4CPUs.  However, the qsub command that I'm using seems to send all jobs
to only one node of the cluster, and then they run sequentially.  I
figured as a test I could start with a script with 10 lines like:

qsub  -N temp1 -q workq  /tmp/launch.csh

I've try adding in resource parameters, but it hasn't helped.  Is there
something obvious that I'm missing.  This would have worked on an LSF
based cluster.  I know that MPI jobs will run as the cluster is
currently configured.

Thanks again.


--
----
James Fenton
LION bioscience, Inc.
[EMAIL PROTECTED]
phone:  617-245-5421
FAX:    617-245-5401




-------------------------------------------------------
This SF.NET email is sponsored by:
SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
http://www.vasoftware.com
_______________________________________________
Oscar-users mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/oscar-users

Reply via email to