From: Jim Phillips (jim_at_ks.uiuc.edu)
Date: Thu Nov 13 2003 - 15:55:45 CST
Hi,
It sounds like your job is being killed and restarted rather than
suspended. I don't understand how PBS would be able to suspend NAMD.
-Jim
On Mon, 10 Nov 2003, J. Rui Rodrigues wrote:
> Hi,
>
> I am not sure if this is a namd question...
>
> I am using NAMD 2.5 for Linux-i686 in a Pentium cluster running PBS.
> This cluster has 2 execution queues for short and long jobs, the "short"
> queue having priority over the "long" queue. When a short job is submitted
> and there are no availables nodes, a job on the long queue is suspended until
> the short job is completed.
>
> My problem is that my namd runs are sent to the long queue and, after being
> suspended, the job restarts from the beginning and not at the point it was
> suspended.
>
> Any idea on how to solve this?
>
> Here is my submission script:
> #!/bin/bash
> #PBS -q centopeia_i_at_pata38.fisica
> #PBS -l nodes=8
> #PBS -l cput=100:00:00
> #PBS -N NAMD
> #PBS -o NAMD_prod.out
> #PBS -e NAMD_prod.err
> cd /home/biomd/WT
> echo "group main" > mynodes
> awk '{print " host " $0}' < $PBS_NODEFILE >> mynodes
>
> /home/biomd/bin/charmrun ++verbose ++p 8 ++nodelist
> mynodes /home/biomd/bin/namd2 /home/biomd/WT/prod.namd
> >> /home/biomd/WT/prod.log
>
> Thanks,
> Rui Rodrigues
>
> --
> J. Rui Rodrigues
> Departamento de Química
> Faculdade de Ciências e Tecnologia
> Universidade de Coimbra
> 3004-535 Coimbra
> Portugal
>
>
>
This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:37:08 CST