> > Hi Christoph! > > It's all fine only if you are not running out of the max wall-time. > In many cases that max is 24 hours, which is way too small for any serious > system. > > So, the question must be, how to arrange running votca on an HPC cluster with > self-resubmitting ability, for both/either votca and/or the simulation engine? > > Cheers > > Andrey
Here is Christoph's response for those interested: ===== Concerning your question: 1.) set cg.inverse.dlpoly.command to something like "mpirun -np 8 DLPOLY.Z" to make dlpoly run using mpi 2.) do NOT run csg_inverse using mpirun in the queueing script 3.) use csg_inverse --wall-time $((24*3600)) to give the wall clock time (in sec) to csg_inverse. csg_inverse will stop if it cannot make another iteration step. 4.) if you want automatic resubmission add a line like [[ $? -eq 0 && ! -f done ]] && qsub $QSCRIPT to the end of your submit script. Christoph ===== -- You received this message because you are subscribed to the Google Groups "votca" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/votca. For more options, visit https://groups.google.com/groups/opt_out.
