Hi Edward, Wanted to update you a prior problem I had. So I ran relax with multiple processors and single processor with the same data set, and obtained identical data on my computer. So the previous issue I had with missing data I don't think is related to relax itself. Just wanted to let you know.
Sincerely, Sam On Thu, Oct 6, 2016 at 3:05 PM, Mahdi, Sam <sam.mahdi....@my.csun.edu> wrote: > Hi Edward, > > I was talking to the developer of mpi4py, and he said that mpi4py should > work with both, even if you have both of them installed, by loading the > appropriate module (openmpi or mpich) I should be able to run mpi4py. He > said he suspects there was probably a problem with the installation of > openmpi itself (improper compilation, or the package itself was improperly > configured itself). Regardless, it finally works Thanks to both of you for > your help with relax and my multi-processor problem. > > Sincerely, > Sam > > On Thu, Oct 6, 2016 at 2:39 PM, Edward d'Auvergne <edw...@nmr-relax.com> > wrote: > >> On Thursday, 6 October 2016, Mahdi, Sam <sam.mahdi....@my.csun.edu> >> wrote: >> >>> Hi Troels and Edward, >>> >>> I was able to get relax working on a multi-processor platform finally. I >>> tried using the mpich instead of openmpi. >>> module load mpi/mpich-x86_64 >>> mpirun -np 7 ./relax --multi='mpi4py' >>> >>> and it loaded up relax with 1 master and 7 slaves. I don't exactly >>> understand why openmpi doesn't work. I have gotten both from the package >>> list, and installed them directly., and I have mpi4py for both mpich and >>> openmpi (I also got these from the package list and installed directly), >>> but this seemed to finally get it to work. >>> >> >> Hi Sam, >> >> This sounds like a classic software clash, with OpenMPI loosing out to >> MPICH with respect to mpi4py. Maybe uninstalling all MPI packages and just >> having openmpi and python-mpi4py would have solved this. Anyway, as long >> as something works, that's all that is needed. Thank you Troels for all of >> your help! >> >> Regards, >> >> Edward >> > > _______________________________________________ relax (http://www.nmr-relax.com) This is the relax-users mailing list relax-users@gna.org To unsubscribe from this list, get a password reminder, or change your subscription options, visit the list information page at https://mail.gna.org/listinfo/relax-users
