[Wien] Error in parallel execution

Stefaan Cottenier stefaan.cottenier at ugent.be
Tue Jul 27 15:36:55 CEST 2010


> Could there be any known
> (but obscure) reason why lapw1para would not be able to write to its files,
> but lapw0para would?

With k-point parallelization, lapw0 runs serially on the machine where  
you have started the job. It does not need ssh. Only lapw1 and lapw2  
are run parallelly.

In lapw1para, set debug to 1 to get more info in case.dayfile:

set debug       = 1             # verbosity of debugging output

Execute eventually this line

$remote $machine[$p] "cd $PWD;$t $exe ${def}_$loop.def ;fixerror_lapw  
${def}_$loop";rm -f .lock_$lockfile[$p]) >& .stdout1_$loop; sortoutput  
1_$loop &

from lapw1para on the command line (with appropriate values for the  
variables and with p=1, i.e. basically 'lapw1 lapw1_1.def') to see  
interactively where it goes wrong.

Stefaan





More information about the Wien mailing list