Problems with mpi version of antibody.linuxgccrelease

Member Site Forums Rosetta 3 Rosetta 3 – Applications Problems with mpi version of antibody.linuxgccrelease

Viewing 0 reply threads
  • Author
    Posts
    • #3443
      Anonymous

        Dear all, 

        When I run the “antibody.mpi.linuxgccrelease”, like

             antibody.mpi.linuxgccrelease -fasta vhh.fasta -vhh_only | tee grafting.log

        I always get the following strange information when the program approachs finishing and all grafted models have been produced:

        “””mpirun has exited due to process rank 5 with PID 30003 on

        node cn10386 exiting improperly. There are three reasons this could occur:

        1. this process did not call “init” before exiting, but others in

        the job did. This can cause a job to hang indefinitely while it waits

        for all processes to call “init”. By rule, if one process calls “init”,

        then ALL processes must call “init” prior to termination.

        2. this process called “init”, but exited without calling “finalize”.

        By rule, all processes that call “init” MUST call “finalize” prior to

        exiting or it will be considered an “abnormal termination”

        3. this process called “MPI_Abort” or “orte_abort” and the mca parameter

        orte_create_session_dirs is set to false. In this case, the run-time cannot

        detect that the abort call was an abnormal termination. Hence, the only

        error message you will receive is this one.

        This may have caused other processes in the application to be

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

        You can avoid this message by specifying -quiet on the mpirun command line.”””

        Do you know why does this information appear? should I add the “-quiet” as it suggests?

        Thank you in advance.

    Viewing 0 reply threads
    • You must be logged in to reply to this topic.