[aspect-devel] Problem running ASPECT on Stampede2

[FELIPE ORELLANA ROVIROSA]

  Dear all,

     Thanks a lot for your prompt comments.

     Diogo you got the full idea of the problem !

    All what caused the problem was my parameter file, it was the 3D
convection test example from previous ASPECT versions.

    I downloaded the latest prm for the case, and it runs, it produces
output. All good.

    Indeed, in this newest ASPECT versions, there is a separate subsection
for the solver parameters..

    My ASPECT 2.0.1, using deal.ii 9.0.0, trilinos 12.10.1, p4est 2.0.0
seems to run correctly on Stampede2.

  Thank you Diogo, thank you guys, thank you very much.

cheers,
Felipe



On Fri, Jun 29, 2018 at 8:51 AM, Diogo Louro Lourenco <dlourenco at ucdavis.edu
> wrote:

> I believe that in previous Aspect versions "Linear solver tolerance" would
> go into the first section of the .prm file (together with Dimension, Output
> directory, etc), while in Aspec 2.0 it goes into:
>
> subsection Solver parameters
>   subsection Stokes solver parameters
>     set Linear solver tolerance = ...
>   end
> end
>
> Maybe it is an incompatibility between the Aspect version and the .prm
> file you are using?
>
> Cheers,
> Diogo
>
> On Fri, Jun 29, 2018 at 2:51 AM Max Rudolph <maxrudolph at ucdavis.edu>
> wrote:
>
>> Felipe,
>> The error message says that you tried to set the parameter 'Linear solver
>> tolerance' in the wrong subsection of your input file. Maybe there is an
>> 'end' statement missing somewhere in the prm file.
>>
>> Max
>>
>> On Thu, Jun 28, 2018 at 5:31 PM FELIPE ORELLANA ROVIROSA <
>> f_orellana at berkeley.edu> wrote:
>>
>>>
>>>  Hi all,
>>>
>>>        I have installed ASPECT on Stampede2.
>>>
>>>        I am getting error when running the test example parameter file
>>> for 3D convection.
>>>
>>>        I am pasting the error message at the end. It fails on not
>>> getting the instruction for linear solver tolerance. At first I was using
>>> 16 nodes, 1088 cores. Then I used 1 node, 68 cores.
>>>
>>>     Do you guys have a clue on what is happening?
>>>
>>>     Should I compile the debug version?
>>>
>>> cheers,
>>> Felipe
>>>
>>> ibrun /work/04020/unfelipe/software/candi/aspect/aspect exAsp.prm
>>> TACC:  Starting up job 1692989
>>> TACC:  Starting parallel tasks...
>>> ------------------------------------------------------------
>>> -----------------
>>> -- This is ASPECT, the Advanced Solver for Problems in Earth's
>>> ConvecTion.
>>> --     . version 2.0.1 (aspect-2.0, 2863594)
>>> --     . using deal.II 9.0.0
>>> --     . using Trilinos 12.10.1
>>> --     . using p4est 2.0.0
>>> --     . running in OPTIMIZED mode
>>> --     . running with 68 MPI processes
>>> -- How to cite ASPECT: https://aspect.geodynamics.org/cite.html
>>> ------------------------------------------------------------
>>> -----------------
>>>
>>>     Line <18> of file <input string>: No entry with name <Linear solver
>>> tolerance> was declared in the current subsection.
>>>
>>> ERROR: Uncaught exception in MPI_InitFinalize on proc 0. Skipping
>>> MPI_Finalize() to avoid a deadlock.
>>> ERROR: Uncaught exception in MPI_InitFinalize on proc 3. Skipping
>>> MPI_Finalize() to avoid a deadlock.ERROR: Uncaught exception in
>>> MPI_InitFinalize on proc 8. Skipping MPI_Finalize() to avoid a
>>> deadlock.ERROR: Uncaught exception in MPI_InitFinalize on proc 12. Skipping
>>> MPI_Finalize() to avoid a deadlock.
>>> ERROR: Uncaught exception in MPI_InitFinalize on proc 16. Skipping
>>> MPI_Finalize() to avoid a deadlock.
>>> ERROR: Uncaught exception in MPI_InitFinalize on proc 34. Skipping
>>> MPI_Finalize() to avoid a deadlock.
>>> ERROR: Uncaught exception in MPI_InitFinalize on proc 55. Skipping
>>> MPI_Finalize() to avoid a deadlock.
>>> ERROR: Uncaught exception in MPI_InitFinalize on proc 58. Skipping
>>> MPI_Finalize() to avoid a deadlock.
>>> ERROR: Uncaught exception in MPI_InitFinalize on proc 62. Skipping
>>> MPI_Finalize() to avoid a deadlock.
>>> ERROR: Uncaught exception in MPI_InitFinalize on proc 65. Skipping
>>> MPI_Finalize() to avoid a deadlock.
>>> ERROR: Uncaught exception in MPI_InitFinalize on proc 66. Skipping
>>> MPI_Finalize() to avoid a deadlock.
>>>
>>>
>>> ----------------------------------------------------
>>> Exception 'dealii::ExcMessage ("Invalid input parameter file.")' on rank
>>> 0 on processing:
>>> ERROR: Uncaught exception in MPI_InitFinalize on proc 1. Skipping
>>> MPI_Finalize() to avoid a deadlock.
>>>
>>> ERROR: Uncaught exception in MPI_InitFinalize on proc 4. Skipping
>>> MPI_Finalize() to avoid a deadlock.
>>>
>>> _______________________________________________
>>> Aspect-devel mailing list
>>> Aspect-devel at geodynamics.org
>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/aspect-devel
>>
>> _______________________________________________
>> Aspect-devel mailing list
>> Aspect-devel at geodynamics.org
>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/aspect-devel
>
>
> _______________________________________________
> Aspect-devel mailing list
> Aspect-devel at geodynamics.org
> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/aspect-devel
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.geodynamics.org/pipermail/aspect-devel/attachments/20180629/4273ea54/attachment.html>