[Wien] Parallel version Lapw1 crashed

Bo Qiu 200210qb at gmail.com
Thu Mar 27 22:55:13 CET 2008


Thanks for your reply! It works! Just one more question, I tried to use
w2web to run parallel calculations, and in the Configuration, I filled in:

pbs=cat ~/head > w2webjob;cat %f >> w2webjob;chmod +x w2webjob;qsub w2webjob

for execution types, the resulting w2webjob will be something like:

#PBS -l nodes=4,cput=24:00:00,walltime=24:00:00
#PBS -me -M 200210qb at gmail.com
#!/bin/sh
date
cd /home/nanoenergy/a/qiub/four
run_lapw -p  -ec 0.0001 -cc 0.001 -NI >/home/nanoenergy/a/qiub/four/STDOUT
2>&1

rm /home/nanoenergy/a/qiub/four/.command.333364.25124

However, I'm using tcsh shell, when I run this jobscript it will complain
that "Ambiguous output redirect." seemingly due to different syntax between
bash and tcsh. If I remove ">/home/nanoenergy/a/qiub/four/STDOUT 2>&1"
manually then it can run but I cannot get the STDOUT file. Just want to ask
whether if it is possible to fix this problem by modifying the lines
describing execution types. Thanks a lot!

Sincerely,
Bo

On Thu, Mar 27, 2008 at 2:31 PM, Stefaan Cottenier <
stefaan.cottenier at fys.kuleuven.be> wrote:

>
> You probably defined your scratch variable such that on your first
> machine it contains
>
> /home/nanoenergy_1/a/qiub/scratch/
>
> and on your second machine
>
> /home/nanoenergy_2/a/qiub/scratch/
>
> etc. And these directories are not mounted on all machines in your
> setup. It is easier to define SCRATCH to be the same on all machines,
> e.g. put it to /tmp. What comes in $SCRATCH does not need to be known by
> other machines.
>
> Stefaan
>
>
> Bo Qiu wrote:
> > Dear Users and Developlers,
> >
> > A few days ago I compiled series version wien2k_08 on a dual core Xeon
> > em64t cluster with tcsh shell successfully, however, when I tried to
> > run parallel version, it cannot work if I use w2web to run parallel
> > calculations automatically.  So I tried to run it directly in the case
> > folder. What I did is just:
> >
> > cd ~/case
> > run_lapw -p -ec 0.0001
> >
> > It ran for a few seconds and I read from the screen:
> >
> >  LAPW0 END
> > LAPW1 - Error
> > LAPW1 - Error
> > LAPW1 - Error
> > LAPW1 - Error
> > LAPW1 - Error
> >
> > >   stop error
> >
> > then I checked the lapw1_1.error:
> >
> >  'INILPW' - can't open unit:
> > 10
> >  'INILPW' -        filename:
> > /home/nanoenergy_1/a/qiub/scratch/four.vector_1
> >  'INILPW' -          status: unknown      form:
> > unformatted
> >  'LAPW1' - INILPW aborted unsuccessfully.
> >
> > for lapw1_2.error to lapw1_5.error, they are basically the same
> >
> > 'INILPW' - can't open unit:
> > 10
> >  'INILPW' -        filename:
> > /home/nanoenergy_5/a/qiub/scratch/four.vector_5
> >  'INILPW' -          status: unknown      form:
> > unformatted
> >  'LAPW1' - INILPW aborted unsuccessfully.
> >
> > And I checked and found there was no folder named as nanoenergy_5 in
> > /home/ folder and there was nothing produced in /scratch.
> >
> > Could you please help me with this? Thanks!
> >
> > Sincerely,
> > Bo
> > ------------------------------------------------------------------------
> >
> > _______________________________________________
> > Wien mailing list
> > Wien at zeus.theochem.tuwien.ac.at
> > http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> >
>
>
> Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>



-- 
Thanks and Best Wishes.

Sincerely yours,
Bo Qiu
http://web.ics.purdue.edu/~qiub/homepage/<http://web.ics.purdue.edu/%7Eqiub/homepage/>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20080327/eddc8b9d/attachment.html


More information about the Wien mailing list