open TELEMAC-MASCARET The mathematically superior suite of solvers

I there, I managed to go through the partitionning of T3DBND by modifying its SUBMIT key in telemac3d.dico. Could you try it to see if it fixes the issue for you guys too? The idea is that in sources/telemac3d/telemac3d.dico, you have to modify the line:
SUBMIT = 'T3DBND-READ;T3DBND;FACUL;BIN;LIT;SELAFIN'
(it may be slightly different for you if you are on an older version than v7p3), into:
SUBMIT = 'T3DBND-READ;T3DBND;FACUL;BIN;LIT;PARAL'

Please let me know about the outcome...
Cheers,
Agnès

Hi Agnès,

Thank you for this fix !
The T3DBND file is not partitioned but only duplicated as many times as there are processors, but I think it's normal as it's the aim of the addition of the SUBMIT key "PARAL".
The simulation goes then further but there is still a problem at the first iteration during the call to write_mesh, more precisely in the cpstvc.f subroutine.


Laurent

I saw this issue too, but I think it's another problem... Is the simulation running in sequential mode?

Cheers,
Agnès

With the v7p2r1, it runs for both sequential and parallel modes (I didn't check the results yet).
With the v7p3r0, the error in write_mesh occurs in both sequential and parallel modes.
It seems there is something broken between the 2 versions..

Cheers,
Laurent

Hi Laurent
We have also encountered this problem but I can't get access to your log file unfortunately as I don't have permission to open it.
Could you upload as a text file with read access please?

Regards
John

Hi,
Here is the log file !


With the fix provided by Agnès, it works with v7p2r1.

Regards,
Laurent

Hi,

I have tried the fix in version v7p2r2 and it is not working for me. I am getting an error in cpstvc.f subroutine, so I would like to see your log file as well in case I am having the same problem as you are in version v7p3r0.

Laurent, I do not have permission to download any of your files either...

Cheers,
Luz

Hi,
I don't know why you can't read my log file... but I think it's the same issue. I didn't have time to look at the differences between the 2 versions, but you can try to run your case with v7p2r1.

================================================================================
ITERATION 0 TIME 0 D 0 H 0 MN 0.0000 S ( 0.0000 S)
================================================================================
APPEL DE PRERES_TELEMAC3D
RETOUR DE PRERES_TELEMAC3D
APPEL DE WRITE_HEADER
RETOUR DE WRITE_HEADER
APPEL DE WRITE_MESH
CPSTVC : FORBIDDEN CASE FOR X AND Y:
X=X SIZE : 59420
Y=TR2D1 MAXIMUM SIZE: 5942
application called MPI_Abort(MPI_COMM_WORLD, 2) - process 0

PLANTE: PROGRAM STOPPED AFTER AN ERROR
RETURNING EXIT CODE: 2
application called MPI_Abort(MPI_COMM_WORLD, 2) - process 1

===================================================================================
= BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
= PID 30989 RUNNING AT moorea
= EXIT CODE: 2
= CLEANING UP REMAINING PROCESSES
= YOU CAN IGNORE THE BELOW CLEANUP MESSAGES
===================================================================================
_____________
runcode::main:
:
|runCode: Fail to run
|$MPIHOME/bin/mpirun -machinefile MPI_HOSTFILE -np 2 /home5/leballeur/tmp_binbnd/forum_files_luz/DartEstuary_3D_bathy_v0.cas_2018-02-23-10h10min34s/out_telemac3d
|~~~~~~~~~~~~~~~~~~
|application called MPI_Abort(MPI_COMM_WORLD, 2) - process 1
|~~~~~~~~~~~~~~~~~~

Luz,

I think it should now work in spherical coordinates, as the error is only related to writing of mesh in lat/lon coordinates.
Let me know if you have troubles.

Hope it helps,
Laurent

Hi Laurent,

So your case now works? Do you have any problems of noisy behaviour close to the boundaries or does it work fine?

Cheers,
Agnès