foamToVTK -parallel fails with processor-cyclic

On tutorials/incompressible/pimpleFoam/LES/periodicHill/steadyState foamToVTK fails. Might be to do with cellZones.

VTK mesh topology: 2.51 s, 640008 kB
Time: 1
    volScalar    : nuTilda nut p
    volVector    : U
    Internal  : "VTK/steadyState_1/internal.vtu"
    Boundary  : "VTK/steadyState_1/boundary/inlet.vtp"
    Boundary  : "VTK/steadyState_1/boundary/outlet.vtp"
    Boundary  : "VTK/steadyState_1/boundary/top.vtp"
    Boundary  : "VTK/steadyState_1/boundary/hills.vtp"
[preston:23831] *** An error occurred in MPI_Recv
[preston:23831] *** reported by process [140731114651649,9]
[preston:23831] *** on communicator MPI_COMM_WORLD
[preston:23831] *** MPI_ERR_TRUNCATE: message truncated
[preston:23831] *** MPI_ERRORS_ARE_FATAL (processes in this communicator will now abort,
[preston:23831] ***    and potentially your MPI job)
    Boundary  : "VTK/steadyState_1/boundary/front.vtp"
    Boundary  : "VTK/steadyState_1/boundary/back.vtp"
[preston:23820] 4 more processes have sent help message help-mpi-errors.txt / mpi_errors_are_fatal
[preston:23820] Set MCA parameter "orte_base_help_aggregate" to 0 to see all help / error messages

Or processorCyclic instead of processor? On processor0 "back" is followed by a "processor", but on processor9 (for example) "back" is followed by a processorCyclic. Need to change nNonProcessor counting in polyBoundaryMesh.

but processorCyclicPolyPatch isA processorPolyPatch

correct - solved elsewhere. The foamToVTK had its own check, since we may wish to write out the processor boundaries from inside the processorN/ directory. fixing.

mentioned in commit ef7dadede94ce40f6be98b573de65c1f18ba61d0

Thanks!

closed

changed title from foamToVTK -parallel with cellZones fails to foamToVTK -parallel fails with processor-cyclic