[petsc-users] Parallel DM example for FEM

Matthew Knepley knepley at gmail.com
Mon Nov 11 06:49:54 CST 2013 

On Mon, Nov 11, 2013 at 6:47 AM, Anthony Vergottis <a.vergottis at ucl.ac.uk>wrote:

> I would also like to ask as I have some confusion. Are the DM objects
> capable of handling all aspects of the parallel computations within a given
> problem?
>
> Let me elaborate - for example if I am passing a mesh structure
> (triangular 3 node elements) as the following:
>
> Elem 0: N1 N5 N6
> etc.....
> where N# is the node tag.
>
> into a DM object and then perform a decomposition for as many processes I
> required. When the mesh partitions are distributed to their processes do I
> then have the ability by using DM objects or something else built into
> PETSc to compute specific calculations on each process without the need to
> hardcode any MPI?
>

I believe the answer is yes. I could be more specific with a concrete
example. For instance, the DM can
calculate a parallel residual without the user ever seeing MPI.

   Matt


> I hope I have explained this well.
>
> Thanks again.
> Adoni
>
>
> On 11 November 2013 12:33, Anthony Vergottis <a.vergottis at ucl.ac.uk>wrote:
>
>> Thanks mate that worked!
>>
>> Thanks.
>> Adoni
>>
>>
>>
>> On 11 November 2013 12:29, Matthew Knepley <knepley at gmail.com> wrote:
>>
>>> On Mon, Nov 11, 2013 at 6:25 AM, Anthony Vergottis <
>>> a.vergottis at ucl.ac.uk> wrote:
>>>
>>>> I compile ex12 but when I run the program I get this in the terminal.
>>>>
>>>
>>> I guess I should make better defaults. Here are the runs I use for ex12:
>>>
>>>
>>> https://bitbucket.org/petsc/petsc/src/a965ca046084fa53248a41da989a0a62cb6266ea/config/builder.py?at=master#cl-180
>>>
>>> You can see that the problem here is that the -petscspace_order has not
>>> been set. So to test P_1 I use
>>>
>>>   -run_type test -refinement_limit 0.0 -bc_type dirichlet -interpolate
>>> 0 -petscspace_order 1 -show_initial -dm_plex_print_fem 1
>>>
>>> and then take away run_type (or use full) to solve the problem. SNES
>>> ex62 is the Stokes problem, and run parameters are
>>> in the same place.
>>>
>>>   Thanks,
>>>
>>>      Matt
>>>
>>>
>>>> adoni at Adoni:~/Desktop/petsc-lat/src/snes/examples/tutorials$ make ex12
>>>> /home/adoni/Desktop/petsc-lat/arch-linux2-c-debug/bin/mpicxx -o ex12.o
>>>> -c -Wall -Wwrite-strings -Wno-strict-aliasing -Wno-unknown-pragmas -g
>>>>  -fPIC    -I/home/adoni/Desktop/petsc-lat/include
>>>> -I/home/adoni/Desktop/petsc-lat/arch-linux2-c-debug/include
>>>>  /home/adoni/Desktop/petsc-lat/src/snes/examples/tutorials/ex12.c
>>>> /home/adoni/Desktop/petsc-lat/arch-linux2-c-debug/bin/mpicxx -Wall
>>>> -Wwrite-strings -Wno-strict-aliasing -Wno-unknown-pragmas -g  -o ex12
>>>> ex12.o -Wl,-rpath,/home/adoni/Desktop/petsc-lat/arch-linux2-c-debug/lib
>>>> -L/home/adoni/Desktop/petsc-lat/arch-linux2-c-debug/lib  -lpetsc
>>>> -Wl,-rpath,/home/adoni/Desktop/petsc-lat/arch-linux2-c-debug/lib -lflapack
>>>> -lfblas -ltriangle -lX11 -lparmetis -lmetis -lpthread -lchaco -lctetgen -lm
>>>> -Wl,-rpath,/usr/lib/gcc/x86_64-linux-gnu/4.6
>>>> -L/usr/lib/gcc/x86_64-linux-gnu/4.6 -Wl,-rpath,/usr/lib/x86_64-linux-gnu
>>>> -L/usr/lib/x86_64-linux-gnu -Wl,-rpath,/lib/x86_64-linux-gnu
>>>> -L/lib/x86_64-linux-gnu -lmpichf90 -lgfortran -lm -lgfortran -lm -lquadmath
>>>> -lm -lmpichcxx -lstdc++ -ldl -lmpich -lopa -lmpl -lrt -lpthread -lgcc_s
>>>> -ldl
>>>> /bin/rm -f ex12.o
>>>> adoni at Adoni:~/Desktop/petsc-lat/src/snes/examples/tutorials$ mpiexec
>>>> -np 2 ./ex12
>>>> [0]PETSC ERROR: --------------------- Error Message
>>>> ------------------------------------
>>>> [0]PETSC ERROR: Invalid argument!
>>>> [0]PETSC ERROR: The section cell closure size 0 != dual space dimension
>>>> 1!
>>>> [0]PETSC ERROR:
>>>> ------------------------------------------------------------------------
>>>> [0]PETSC ERROR: Petsc Development GIT revision: unknown  GIT Date:
>>>> unknown
>>>> [0]PETSC ERROR: See docs/changes/index.html for recent updates.
>>>> [0]PETSC ERROR: See docs/faq.html for hints about trouble shooting.
>>>> [0]PETSC ERROR: See docs/index.html for manual pages.
>>>> [0]PETSC ERROR:
>>>> ------------------------------------------------------------------------
>>>> [0]PETSC ERROR: ./ex12 on a arch-linux2-c-debug named Adoni by adoni
>>>> Mon Nov 11 12:21:49 2013
>>>> [0]PETSC ERROR: Libraries linked from
>>>> /home/adoni/Desktop/petsc-lat/arch-linux2-c-debug/lib
>>>> [0]PETSC ERROR: Configure run at Fri Nov  8 20:03:26 2013
>>>> [0]PETSC ERROR: Configure options --with-clanguage=c++
>>>> --with-shared-libraries=1 --download-f-blas-lapack --download-mpich
>>>> --download-boost --download-triangle --download-ctetgen --download-chaco
>>>> --download-metis --download-parmetis
>>>> [0]PETSC ERROR:
>>>> ------------------------------------------------------------------------
>>>> [0]PETSC ERROR: DMPlexProjectFunctionLocal() line 207 in
>>>> src/dm/impls/plex/plexfem.c
>>>> [0]PETSC ERROR: DMPlexProjectFunction() line 258 in
>>>> src/dm/impls/plex/plexfem.c
>>>> [0]PETSC ERROR: main() line 686 in
>>>> /home/adoni/Desktop/petsc-lat/src/snes/examples/tutorials/ex12.c
>>>> [0]PETSC ERROR: --------------------- Error Message
>>>> ------------------------------------
>>>> [0]PETSC ERROR: Invalid argument!
>>>> [0]PETSC ERROR: The section cell closure size 0 != dual space dimension
>>>> 1!
>>>> [0]PETSC ERROR:
>>>> ------------------------------------------------------------------------
>>>> [0]PETSC ERROR: Petsc Development GIT revision: unknown  GIT Date:
>>>> unknown
>>>> [0]PETSC ERROR: See docs/changes/index.html for recent updates.
>>>> [0]PETSC ERROR: See docs/faq.html for hints about trouble shooting.
>>>> [0]PETSC ERROR: See docs/index.html for manual pages.
>>>> [0]PETSC ERROR:
>>>> ------------------------------------------------------------------------
>>>> [0]PETSC ERROR: ./ex12 on a arch-linux2-c-debug named Adoni by adoni
>>>> Mon Nov 11 12:21:49 2013
>>>> [0]PETSC ERROR: Libraries linked from
>>>> /home/adoni/Desktop/petsc-lat/arch-linux2-c-debug/lib
>>>> [0]PETSC ERROR: Configure run at Fri Nov  8 20:03:26 2013
>>>> [0]PETSC ERROR: Configure options --with-clanguage=c++
>>>> --with-shared-libraries=1 --download-f-blas-lapack --download-mpich
>>>> --download-boost --download-triangle --download-ctetgen --download-chaco
>>>> --download-metis --download-parmetis
>>>> [0]PETSC ERROR:
>>>> ------------------------------------------------------------------------
>>>> [0]PETSC ERROR: DMPlexProjectFunctionLocal() line 207 in
>>>> src/dm/impls/plex/plexfem.c
>>>> [0]PETSC ERROR: DMPlexProjectFunction() line 258 in
>>>> src/dm/impls/plex/plexfem.c
>>>> [0]PETSC ERROR: main() line 686 in
>>>> /home/adoni/Desktop/petsc-lat/src/snes/examples/tutorials/ex12.c
>>>> application called MPI_Abort(MPI_COMM_WORLD, 62) - process 0
>>>> [unset]: aborting job:
>>>> application called MPI_Abort(MPI_COMM_WORLD, 62) - process 0
>>>> application called MPI_Abort(MPI_COMM_WORLD, 62) - process 0
>>>> [unset]: aborting job:
>>>> application called MPI_Abort(MPI_COMM_WORLD, 62) - process 0
>>>>
>>>> This occurs for all process counts.
>>>>
>>>> Thanks.
>>>>
>>>> Adoni
>>>>
>>>>
>>>> On 11 November 2013 12:19, Matthew Knepley <knepley at gmail.com> wrote:
>>>>
>>>>> On Mon, Nov 11, 2013 at 6:10 AM, Anthony Vergottis <
>>>>> a.vergottis at ucl.ac.uk> wrote:
>>>>>
>>>>>> Dear All,
>>>>>>
>>>>>> I am trying to find a good example within PETSc that examples with
>>>>>> some detail the use of DM objects within a parallel and FEM framework. I
>>>>>> have looked at most of the example and have not found anything like this.
>>>>>> Is there any other material available on the subject from any other source
>>>>>> or maybe I missed something within the PETSc full download file?
>>>>>>
>>>>>
>>>>> Did you look at SNES ex12? Is it not understandable?
>>>>>
>>>>>   Thanks,
>>>>>
>>>>>      Matt
>>>>>
>>>>>
>>>>>> Thanks in advance.
>>>>>>
>>>>>> Regards,
>>>>>> Adonis
>>>>>>
>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> What most experimenters take for granted before they begin their
>>>>> experiments is infinitely more interesting than any results to which their
>>>>> experiments lead.
>>>>> -- Norbert Wiener
>>>>>
>>>>
>>>>
>>>
>>>
>>> --
>>> What most experimenters take for granted before they begin their
>>> experiments is infinitely more interesting than any results to which their
>>> experiments lead.
>>> -- Norbert Wiener
>>>
>>
>>
>


-- 
What most experimenters take for granted before they begin their
experiments is infinitely more interesting than any results to which their
experiments lead.
-- Norbert Wiener
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-users/attachments/20131111/c13935b0/attachment-0001.html>

More information about the petsc-users mailing list