[EternalSeekerX]

Hello everyone,

Does anyone know what this error means?

localhost:09256] *** Process received signal ***
[localhost:09256] Signal: Segmentation fault (11)
[localhost:09256] Signal code:  (-6)
[localhost:09256] Failing at address: 0x3e800002428
ExecutionTime = 1498.22 s  ClockTime = 1561 s

End

[localhost:09253] *** Process received signal ***
[localhost:09253] Signal: Segmentation fault (11)
[localhost:09253] Signal code:  (-6)
[localhost:09253] Failing at address: 0x3e800002425
[localhost:09257] *** Process received signal ***
[localhost:09257] Signal: Segmentation fault (11)
[localhost:09257] Signal code:  (-6)
[localhost:09257] Failing at address: 0x3e800002429
[localhost:09258] *** Process received signal ***
[localhost:09258] Signal: Segmentation fault (11)
[localhost:09258] Signal code:  (-6)
[localhost:09258] Failing at address: 0x3e80000242a
--------------------------------------------------------------------------
mpirun noticed that process rank 0 with PID 9253 on node localhost exited on signal 11 (Segmentation fault).
--------------------------------------------------------------------------

This only happens in 4.1, but not 4.0. I am using CentOS 7 and mpi runs fine in all my openfoam installs except foam-extend-4.1.This is using the mpi 1.8.8 from the third party folder, as foam extend 4.1 doesnt compile with systemmpi for me.

[Bazinga]

I recall that I read that there is a bug in 4.1 when GAMG is used in parallel. Not sure if that is the case here.

Note that at the time of writing this (26/11/2019) these packages include a bug, that makes parallel runs crash with an MPI error if the GAMG solver is used for the pressure equation. For this reason I cannot recommend these packages.

https://openfoamwiki.net/index.php/I...oam-extend-4.1

[EternalSeekerX]

Originally Posted by Bazinga

I recall that I read that there is a bug in 4.1 when GAMG is used in parallel. Not sure if that is the case here.



https://openfoamwiki.net/index.php/I...oam-extend-4.1

It could be, because solids4foam in 4.1 seems to work fine with parallel. Are there any tutorial that doesn't use GAMG I can test?

[Bazinga]

Originally Posted by EternalSeekerX

It could be, because solids4foam in 4.1 seems to work fine with parallel. Are there any tutorial that doesn't use GAMG I can test?

Just change the solver yourself in fvSolution to test.

[EternalSeekerX]

Originally Posted by Bazinga

Just change the solver yourself in fvSolution to test.

So I just tried a MRFSimpleFoam tutorial which uses PCG solver for the variables all work. Seems like I would have to change all the places GAMG is specified in fvSolutions to run parallel for now.