Skip to content
GitLab
Closed
Issue created by Hua Zen@lin

problem with v1806 docker image on Linux OS

The problem is briefly described in the following thread: https://www.cfd-online.com/Forums/openfoam-solving/204881-unable-launch-parafoam.html

there seems to be something wrong with mpirun in the docker image.

  1. bash-4.2$ mpirun --help

  3. It looks like opal_init failed for some reason; your parallel process is

  4. likely to abort. There are many reasons that a parallel process can

  5. fail during opal_init; some of which are due to configuration or

  6. environment problems. This failure appears to be an internal failure;

  7. here's some additional information (which may only be relevant to an

  8. Open MPI developer):

  10. opal_shmem_base_select failed

  11. --> Returned value -1 instead of OPAL_SUCCESS

  12. bash-4.2$ paraview

  13. QStandardPaths: XDG_RUNTIME_DIR not set, defaulting to '/tmp/runtime-linzh'

  15. It looks like opal_init failed for some reason; your parallel process is

  16. likely to abort. There are many reasons that a parallel process can

  17. fail during opal_init; some of which are due to configuration or

  18. environment problems. This failure appears to be an internal failure;

  19. here's some additional information (which may only be relevant to an

  20. Open MPI developer):

  22. opal_shmem_base_select failed

  23. --> Returned value -1 instead of OPAL_SUCCESS

  26. It looks like orte_init failed for some reason; your parallel process is

  27. likely to abort. There are many reasons that a parallel process can

  28. fail during orte_init; some of which are due to configuration or

  29. environment problems. This failure appears to be an internal failure;

  30. here's some additional information (which may only be relevant to an

  31. Open MPI developer):

  33. opal_init failed

  34. --> Returned value Error (-1) instead of ORTE_SUCCESS

  37. It looks like MPI_INIT failed for some reason; your parallel process is

  38. likely to abort. There are many reasons that a parallel process can

  39. fail during MPI_INIT; some of which are due to configuration or environment

  40. problems. This failure appears to be an internal failure; here's some

  41. additional information (which may only be relevant to an Open MPI

  42. developer):

  44. ompi_mpi_init: ompi_rte_init failed

  45. --> Returned "Error" (-1) instead of "Success" (0)

  47. *** An error occurred in MPI_Init

  48. *** on a NULL communicator

  49. *** MPI_ERRORS_ARE_FATAL (processes in this communicator will now abort,

  50. *** and potentially your MPI job)

  51. [ce8b1d83bd53:2271] Local abort before MPI_INIT completed successfully; not able to aggregate error messages, and not able to guarantee that all other processes were killed!

Edited by Hua Zen