On Tue, May 12, 2015 at 1:40 PM, George Tzotzos <gtzotzos.me.com> wrote:
> Thank you Jason,
>
> I was running mmpbsa.py but inadvertently I used the following script:
>
> mpirun -np 12 MMPBSA.py -O -i decomp.in -o FINAL_RESULTS_MMPBSA.dat -do
> FINAL_DECOMP__MMPBSA.dat -sp complex_solv.prmtop -cp subA-subB.prmtop -rp
> subA.prmtop -lp subB.prmtop -y 10frame.nc
>
> The culprit seems to be the ‘mpirun -np 12’. Skipping this produces no
> error.
>
That will do it. Just as an FYI, the above command will launch 12 threads
of MMPBSA.py connected by an MPI communicator. However, as is the case
with all serial programs, MMPBSA.py knows nothing about MPI so the 12
threads each run off and try to do *exactly* the same thing (stepping on
each others' toes in the process, resulting in the error you saw).
HTH,
Jason
--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue May 12 2015 - 11:00:03 PDT
