[torqueusers] -l nodes specification

Martin Siegert siegert at sfu.ca
Wed Sep 13 11:54:45 MDT 2006


Hi,

by now I am convinced that this is a bug:

-l nodes=1:ppn=2+3

works, but

-l nodes=1:ppn=2+4

does not. The total number of nodes is 4 with two processors each.
In order to allow requests like

-l nodes=6

I added

set server resources_available.nodect = 8

however, it appears that torque "forgets" this setting when dealing
with more comlicated nodes specifications.

This once more begs for a cleanup of the -l nodes=... and -l ncpus=...
code (i.e., -l nodes=... should request nodes, whereas -l ncpus=...
requests processors) so that one could request something like

-l nodes=1:ppn=2,ncpus=4

Cheers,
Martin

-- 
Martin Siegert
Head, HPC at SFU
WestGrid Site Lead
Academic Computing Services                phone: (604) 291-4691
Simon Fraser University                    fax:   (604) 291-4242
Burnaby, British Columbia                  email: siegert at sfu.ca
Canada  V5A 1S6

On Wed, Sep 13, 2006 at 10:53:34AM +0200, bill wrote:
> Martin Siegert a écrit :
> >Hi,
> >
> Hi
> 
> >I am trying to submit a job with the following nodes specification:
> >
> >-l nodes=1:ppn=2+5
> >
> I think this is:
> I want 1 node with 5 ppn, plus I want 5 nodes.

No. This requests one node with two processors plus 5 processors, see
http://www.clusterresources.com/wiki/doku.php?id=torque:2.1_job_submission#nodeexamples

> >(i.e., the first node with two processors and then 5 other processors).
> >I receive an error from qsub:
> >
> >qsub: Job exceeds queue resource limits MSG=cannot locate feasible nodes
> >
> There is not 6 nodes? Perhaps.
> _______________________________________________
> torqueusers mailing list
> torqueusers at supercluster.org
> http://www.supercluster.org/mailman/listinfo/torqueusers


More information about the torqueusers mailing list