Dear Daniel,
I run your script and rerun my own code and then found the origin of the
error. Your parallel code gives the following:
mpirun -np 4 python test.py
mpi4py: processor 1 of 4 :: PyTrilinos: processor 1 of 4 :: FiPy: 6
cells on processor 1 of 4
mpi4py: processor 2 of 4 :: PyTrilinos: processor 2 of 4 :: FiPy: 6
cells on processor 2 of 4
mpi4py: processor 3 of 4 :: PyTrilinos: processor 3 of 4 :: FiPy: 6
cells on processor 3 of 4
mpi4py: processor 0 of 4 :: PyTrilinos: processor 0 of 4 :: FiPy: 4
cells on processor 0 of 4
I had a (lucky) mistake when running my own code because I forgot to add
the --trilinos at the end of the command line. I got a different error
in this case which pointed to the right line in my code (a function
defined over the z coordinate). Once I removed the offending line
everything worked again.
I'm now rewritting the function, probably using .globalValue or similar,
although the gains in speed are very small (something that has already
been discussed here by the way...)
Thank you for your interest
Regards
Marcel
El 16/04/15 a les 19:37, Daniel Wheeler ha escrit:
Hi Marcel,
You might want to try running Trilinos in parallel independently from
FiPy. Just as a further diagnostic. Try running this script
https://github.com/usnistgov/fipy/blob/develop/examples/parallel.py
in parallel.
Also, can you provide the full traceback, that way we can have an idea
about what it is actually doing when it fails.
Cheers,
Daniel
On Thu, Apr 16, 2015 at 9:54 AM, Marcel UJI (IMAP) <[email protected]> wrote:
I'm trying to execute fipy in parallel and found this strange error. Does
anyone know the meaning of it??
I've installed trilinos and could run all the fipy tests without problems,
both in parallel and/or serial. Also my code runs perfectly in serial both
with pysparse and/or trilinos (p.e. mpirun -np 1 code.py --trilinos, etc),
but fails with the above error once I try to run it with 2 or more
processes. Additionally the error message does not seem to depend on the
solver used (I tried several).
My code uses Gmsh3D to compute the mesh and calls a couple of external
fortran programs to compute complex fields on the mesh which are then used
to define the equations. I can give more details if needed.
Marcel
--
Dr. Marcel Aguilella-Arzo
Professor Titular d'Universitat, Física Aplicada
Coordinador de la Subcomissió d'Especialitat de Ciències Experimentals i
Tecnologia
del Màster Universitari en Professor d'Educació Secundària Obligatòria i
Batxillerat,
Formació Professional i Ensenyament d'Idiomes
Departament de Física
Escola Superior de Tecnologia i Ciències Experimentals
Universitat Jaume I
Av. Sos Baynat, s/n
12071 Castelló de la Plana (Spain)
+34 964 728 046
[email protected]
_______________________________________________
fipy mailing list
[email protected]
http://www.ctcms.nist.gov/fipy
[ NIST internal ONLY: https://email.nist.gov/mailman/listinfo/fipy ]
--
Dr. Marcel Aguilella-Arzo
Professor Titular d'Universitat, Física Aplicada
Coordinador de la Subcomissió d'Especialitat de Ciències Experimentals i
Tecnologia
del Màster Universitari en Professor d'Educació Secundària Obligatòria i
Batxillerat,
Formació Professional i Ensenyament d'Idiomes
Departament de Física
Escola Superior de Tecnologia i Ciències Experimentals
Universitat Jaume I
Av. Sos Baynat, s/n
12071 Castelló de la Plana (Spain)
+34 964 728 046
[email protected]
_______________________________________________
fipy mailing list
[email protected]
http://www.ctcms.nist.gov/fipy
[ NIST internal ONLY: https://email.nist.gov/mailman/listinfo/fipy ]
