[gmx-users] gromacs hanging at end of parallel run
David van der Spoel
spoel at xray.bmc.uu.se
Tue Apr 12 14:43:34 CEST 2005
On Tue, 2005-04-12 at 11:46 +0100, J Kendrick wrote:
> I'd be grateful of some pointers as to where to look for this problem.
> I am running gromacs under mpi on an intel cluster.
> The single node version of the program works
> The parallel version (3.2.1) appears to finish the calculation but then
> leaves all of the processes running. The trajectory file is written and a
> summary of the performance on each of the nodes is written, but it doesn't
> finish. This behaviour seems to depend on the calculation and the number of
> I have switched debug on, with scripts to work under gridengine. This
> example (methanol/water from the tutorial) will run on 2 and 4 nodes. At
> one point I thought it was fftw causing the problem so I have compiled and
> linked with both fftw2.1.3 and fftw2.1.5.
> I currently think that the code is in MPI_Finalize and waiting for IO
> completion, but I am not sure what to look at next.
Sounds like MPICH.
There is a workaround in the CVS code, the MPI_Finalize call is replaced
void gmx_finalize(t_commrec *cr)
fprintf(debug,"In gmx_finalize. Will try to synchronize the ring
fprintf(debug,"Succesfully did so! Exiting now.\n");
fprintf(debug,"Will call MPI_Finalize now\n");
ret = MPI_Finalize();
fprintf(debug,"Return code from MPI_Finalize = %d\n",ret);
> Any help gratefully received.
> gmx-users mailing list
> gmx-users at gromacs.org
> Please don't post (un)subscribe requests to the list. Use the
> www interface or send it to gmx-users-request at gromacs.org.
David van der Spoel, PhD, Assoc. Prof., Molecular Biophysics group,
Dept. of Cell and Molecular Biology, Uppsala University.
Husargatan 3, Box 596, 75124 Uppsala, Sweden
phone: 46 18 471 4205 fax: 46 18 511 755
spoel at xray.bmc.uu.se spoel at gromacs.org http://xray.bmc.uu.se/~spoel
More information about the gromacs.org_gmx-users