Issue running burnup calculation

User Support
1

Hello developers

I hope you are all doing well. I am trying to run burnup on a small cluster we have. Normal calculations of OpenMC run fine on it but when i try to run burnup calculation with

mpirun -np 4 -f ~/machines python3 test.py

it throws error shown below. Can you guide in this regard?

Thanks

Fatal error in PMPI_Comm_size: Invalid communicator, error stack:
PMPI_Comm_size(109): MPI_Comm_size(comm=0x3380e570, size=0x7f453694fd1c) failed
PMPI_Comm_size(66).: Invalid communicator
Fatal error in PMPI_Comm_size: Invalid communicator, error stack:
PMPI_Comm_size(109): MPI_Comm_size(comm=0xe05d8570, size=0x7fade3719d1c) failed
PMPI_Comm_size(66).: Invalid communicator
Fatal error in PMPI_Comm_size: Invalid communicator, error stack:
PMPI_Comm_size(109): MPI_Comm_size(comm=0x5e740570, size=0x7f2061881d1c) failed
PMPI_Comm_size(66).: Invalid communicator

2

This kind of error usually occurs when using a mix of MPI implementations during execution. Is it possible that the OpenMC executable on your cluster is built with one version of MPI (MPICH for example) while the version of mpi4py is using another (OpenMPI for example).