Re: NAMD 2.7b1 on NCSA Abe

From: Peter Freddolino (petefred_at_ks.uiuc.edu)
Date: Fri May 01 2009 - 06:41:23 CDT

Jim's run script has a couple of extra lines to start/stop mpd... you
might want to try the following around your mpirun line:

mvapich2-start-mpd
setenv MV2_SRQ_SIZE 4000
setenv VIADEV_SMP_EAGERSIZE 64
setenv VIADEV_SMPI_LENGTH_QUEUE 256
setenv VIADEV_ADAPTIVE_RDMA_LIMIT 0
setenv VIADEV_RENDEZVOUS_THRESHOLD 50000

mpirun -machinefile $PBS_NODEFILE -np $NPROCS
~jphillip/NAMD_2.7b1_Linux-x86_64/namd2 inputs/config_heat.namd >& logfile

mpdallexit

(the mpirun block should be one line but will probably be wrapped in
this email).

Jim, looks like the permissions on the 2.7b1 runbatch file are 750...
any chance we can make this world readable?

Best,
Peter

Gianluca Interlandi wrote:
> Hi Peter,
>
> I cannot use ~jphillip/NAMD_scripts/runbatch_2.7b1 because it's not
> readable or executable for the world:
>
> -rwxr-x--- 1 jphillip tvn 1488 Apr 1 20:08
> /u/ac/jphillip/NAMD_scripts/runbatch_2.7b1
>
> /u/ac/jphillip/NAMD_scripts/runbatch_2.7b1: Permission denied.
>
> and I'm not in group "tvn".
>
> Sorry, I post here the full batchscript:
>
> #!/bin/csh
> #PBS -l nodes=4:ppn=8
> #PBS -l walltime=02:00:00
> #PBS -l mem=4gb
> #PBS -N T300_1auqe
> #PBS -j oe
> #PBS -q normal
>
> echo $PBS_NODEFILE
> cat $PBS_NODEFILE
>
> cd $PBS_O_WORKDIR
>
> # Load necessary software
> soft add +mpichvmi-intel
>
> # Determine the number of CPUs
> set NPROCS = `wc -l $PBS_NODEFILE | awk '{print $1}'`
>
> mpirun -machinefile $PBS_NODEFILE -np $NPROCS
> ~jphillip/NAMD_2.7b1_Linux-x86_64/namd2 inputs/config_heat.namd > logfile
>
>
> Thanks for your help.
>
> Gianluca
>
>
> On Thu, 30 Apr 2009, Peter Freddolino wrote:
>
>> Hi Gianluca,
>> where is $NPROCS defined? I don't see it defined in your script before
>> it is used. Any reason you're not just using the official 2.7b1 script
>> (~jphillip/NAMD_scripts/runbatch_2.7b1)?
>> Peter
>>
>> Gianluca Interlandi wrote:
>>> Hi Peter,
>>>
>>> Sorry, I made a mistake when I copy-pasted the command line into the
>>> e-mail. I am actually using the namd binary:
>>>
>>> mpirun -machinefile $PBS_NODEFILE -np $NPROCS \
>>> ~jphillip/NAMD_2.7b1_Linux-x86_64/namd2 config.namd
>>>
>>> Gianluca
>>>
>>> On Thu, 30 Apr 2009, Peter Freddolino wrote:
>>>
>>>> Hi Gianluca,
>>>> ~jphillip/NAMD_2.7b1_Linux-x86_64 is not the namd binary;
>>>> ~jphillip/NAMD_2.7b1_Linux-x86_64/namd2 is the namd binary. You might
>>>> want to start by making that the command that is being run by mpirun.
>>>> Best,
>>>> Peter
>>>>
>>>> Gianluca Interlandi wrote:
>>>>> I have been trying to run NAMD2.7b1 on Abe but I have run into some
>>>>> problems. I am using:
>>>>>
>>>>> #!/bin/csh
>>>>> #PBS -l nodes=4:ppn=8
>>>>> #PBS -l walltime=02:00:00
>>>>>
>>>>> mpirun -machinefile $PBS_NODEFILE -np $NPROCS \
>>>>> ~jphillip/NAMD_2.7b1_Linux-x86_64 config.namd
>>>>>
>>>>> I found the executable thanks to the NAMD WIKI:
>>>>> http://www.ks.uiuc.edu/Research/namd/wiki/index.cgi?NamdAtNCSA
>>>>>
>>>>> It seems that NAMD instead of running in parallel it starts 32
>>>>> independent single CPU simulations:
>>>>>
>>>>> Charm++> cpu topology info is being gathered!
>>>>> Charm++> 1 unique compute nodes detected!
>>>>>
>>>>> It repeats the above 32 times. Am I using it incorrectly?
>>>>>
>>>>> Thanks,
>>>>>
>>>>> Gianluca
>>>>
>>>>
>>>
>>> -----------------------------------------------------
>>> Gianluca Interlandi, PhD gianluca_at_u.washington.edu
>>> +1 (206) 685 4435
>>> +1 (206) 714 4303
>>> http://artemide.bioeng.washington.edu/
>>>
>>> Postdoc at the Department of Bioengineering
>>> at the University of Washington, Seattle WA U.S.A.
>>> -----------------------------------------------------
>>
>
> -----------------------------------------------------
> Gianluca Interlandi, PhD gianluca_at_u.washington.edu
> +1 (206) 685 4435
> +1 (206) 714 4303
> http://artemide.bioeng.washington.edu/
>
> Postdoc at the Department of Bioengineering
> at the University of Washington, Seattle WA U.S.A.
> -----------------------------------------------------

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:52:43 CST