[Wien] Running k-point parallel:Changs in .machines file

[Ghosh SUDDHASATTWA]

Dear Wien2k Users, 
Please ignore my previous mail 

The machine file was 
1:nx80:2
1:nx80
1:nx80

When I actually started with the command line 

ssh nx80 "cd /group5/.......; run_lapw -cc 0.0001 -ec 0.00005 -p"

The dayfile did show "2 parallel jobs are running" 
But afterwards, lapw 1 crashed. 
Since SGE is not integrated with Wien2k, I would like to know the format to
run with 8 cpu's in one node 
I tried with all the changes as has been suggested by Dr.Marks in some of
his previous mails, but it still gave the same error. 
Any suggestions would be of great help 

Suddhasattwa Ghosh 


-----Original Message-----
From: wien-bounces at zeus.theochem.tuwien.ac.at
[mailto:wien-bounces at zeus.theochem.tuwien.ac.at] On Behalf Of Ghosh
SUDDHASATTWA
Sent: Saturday, April 24, 2010 11:48 AM
To: 'A Mailing list for WIEN2k users'
Subject: Re: [Wien] Running k-point parallel:Changs in .machines file

Dear Wien2k users, 
Thanks for the mail. Right now, I am trying to figure out the way to
integrate all the cpu's in one node That is 8 cpu's. 
I would like to run the code with 8 cpu's .

The machines file was 
1:nx80:8
1:nx80
1:nx80
1:nx80

-----Original Message-----
From: wien-bounces at zeus.theochem.tuwien.ac.at
[mailto:wien-bounces at zeus.theochem.tuwien.ac.at] On Behalf Of
Eric.Duverger at pu-pm.univ-fcomte.fr
Sent: Thursday, April 22, 2010 7:53 PM
To: wien at zeus.theochem.tuwien.ac.at
Subject: Re: [Wien] Running k-point parallel:Changs in .machines file

Dear Wien2k users
for information you could find a sge script running for parallel simulation

++++++++++++++

#!/bin/bash
#
#$ -o $JOB_NAME.o$JOB_ID
#$ -N wien1
#$ -pe impi_tight 51
#$ -p batch
#$ -q normal15d@@nonsusp15d
#$ -V
#$ -l h_rt=60:00:0
#############$ -l mt=23.4G


source /Softs/Intel/Compiler/11.0/069/bin/ifortvars.sh intel64
source /Softs/Intel/Compiler/11.0/069/bin/iccvars.sh intel64
source /Softs/Intel/impi/3.2.1.009/bin64/mpivars.sh
export WIENROOT="/Data/Users/eduverger/wien2k"
export PATH="$WIENROOT:$PATH"
export SCRATCH="/Data/Users/eduverger/wien/SiO2"
script=EW1_run.pbs
cores=1

#$ -V
#$ -cwd
#$ -o Result.out
#$ -e Result.err

# Set internal parallelization code in mkl to only use
# eight thread per process.
export OMP_NUM_THREADS=8

# some information
echo "Got $NSLOTS slots." >> Result.out
echo "Got $NSLOTS slots." >> Result.err

# read the mpi machines files (generated by the sge)
proclist=(`cat $TMPDIR/machines`)
nproc=$NSLOTS
echo $nproc nodes for this job: $proclist

rm .machines

# Convert proclist to one line per slot/k-point.
# In a single queue all nodes have equal performance.
for a in ${proclist[*]}; do
     echo 1:$a >> .machines
done


echo 'granularity:1' >>.machines
echo 'extrafine:1' >>.machines

set i=1
echo -n 'lapw0:'$proclist " " >>.machines
echo ' ' >> .machines


cat $TMPDIR/.machines > toto.txt
NNODES=`cat .machines | uniq | wc -l`
cat .machines | sort | uniq | tee $PWD/mpd.hosts


run_lapw -p -ec 0.0001 -NI
mpdexit


+++++++++++


Best Regards

E Duverger



Quoting "Ghosh SUDDHASATTWA" <ssghosh at igcar.gov.in>:

> Dear Wien2k users,
>
> I am working on a node "nx80" with 8 processors with 1080 k-points
>
> So I made the following changes in the .machines file
>
>
>
> Granularity=1
>
> 1:nx80:1
>
> Residue:nx80:1
>
>
>
> When I submitted the job in the command line with
>
> Run_lapw -cc 0.0001 -ec 0.00005 -p
>
>
>
> It gave an error
>
> Lapw crashed
>
> Can anybody tell me what exactly we have to change in the .machines file.
>
>
>
> Any suggestions would be of great help
>
>
>
> S Ghosh
>
>
>
>
>
>
>
>



----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.
_______________________________________________
Wien mailing list
Wien at zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien

_______________________________________________
Wien mailing list
Wien at zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien