CFD Online Discussion Forums - MPI issues with my own solver

Hi everybody!

I'm not sure if this can be just a running problem, a programming problem or a bug, so I will write here hoping it's ok.

I programmed my own solver ( for joint fluid dynamics and electrical problems ) which is working really fine on serial runs (with around 10^5 mesh cells) but as i now want to use it for a non 2D case, my mesh comes up to 1.5kk cells and I switched to parallel.

I tested before with the breakingDam classic example, which works fine, then i switched to my own stuff

I decomposePar'd the domain - all fine - and then i ran the solver obtaining the following message:

Code:

[moon01-01:19649] *** An error occurred in MPI_Recv

[moon01-01:19649] *** on communicator MPI_COMM_WORLD

[moon01-01:19649] *** MPI_ERR_TRUNCATE: message truncated

[moon01-01:19649] *** MPI_ERRORS_ARE_FATAL (your MPI job will now abort)

                --------------------------------------------------------------------------

mpirun has exited due to process rank 0 with PID 19649 on

node moon01-01 exiting without calling "finalize". This may

have caused other processes in the application to be

terminated by signals sent by mpirun (as reported here).

--------------------------------------------------------------------------

the strange fact is that this error comes up every time at the exact same stage in the solving: I have a newton iteration stage for a nonlinear problem, and this comes up always after the 15th iteration, before the 16th linear solving of the problem.

does anyone have a hint about what's going on?