[Wien] Wien2k mpi execution

[Peter Blaha]

For sure your script does not generate a   "lapw0: host1 host 2 .... " lie,
so lapw0 will run only in sequential mode.

Otherwise, check your .machines file. Was it ok ??? How does it look ?

When with run -p only the seuential versions run, it is usually a problem with the .machine
file, because if a mpi-program is started "in a wrong"  way, it should crash.

What does case.dayfile say ??

Setting for MPI_REMOTE and WIEN_MPIRUN  are crucial and depend on your specific
intallation / mpi version, ......

Am 23.02.2012 21:03, schrieb Celine Hin:
> Dear Wien2k users,
>
> WIEN2k has been compiled with fine grain parallelism support to use MPI. A test job was submitted to use 8 nodes using the queuing system. The job was running only one node
> and the other 7 nodes were idle. I am wondering if there is any error in the .machine file generated by the template qsub-job0_lapw script which has been attached along
> with this email. The installation has generated the following MPI applications
>
> lapw0_mpi lapw1c_mpi lapw1_mpi lapw2c_mpi lapw2_mpi
>
> The contents of parallel_options file is as follows:
>
> setenv USE_REMOTE 1
>
> setenv MPI_REMOTE 1
>
> setenv WIEN_GRANULARITY 1
>
> setenv WIEN_MPIRUN "mpirun -np _NP_ _EXEC_"
>
> Best,
> Celine Hin.
>
>
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien

-- 
-----------------------------------------
Peter Blaha
Inst. Materials Chemistry, TU Vienna
Getreidemarkt 9, A-1060 Vienna, Austria
Tel: +43-1-5880115671
Fax: +43-1-5880115698
email: pblaha at theochem.tuwien.ac.at
-----------------------------------------