Ralph, thanks for the info
can you detail your full mpirun command line, the number of servers you are > using, the btl involved and the ompi release that can be used to reproduce > the issue ? > > > Running on only one server, using the current head of the svn repo. My > cluster only has Ethernet, and I let it freely choose the BTLs (so I > imagine the candidates are sm,self,tcp,vader). The cmd line is really > trivial: > > is MPSS installed and loaded ? if yes, scif is also a candidate > mpirun -n 1 ./loop_spawn > > I modified loop_spawn to only run 100 iterations as I am not patient > enough to wait for 1000, and the number of iters isn't a factor so long as > it is greater than 1. When the parent calls finalize, I get one of the > following emitted for every iteration that was done: > > dpm_base_disconnect_init: error -12 in isend to process 0 > > so we do the same thing but have different behaviour ... just to be sure : are we talking about the loop_spawn test from the ibm test suite available at http://svn.open-mpi.org/svn/ompi-tests/trunk/ibm/dynamic/loop_spawn.c and http://svn.open-mpi.org/svn/ompi-tests/trunk/ibm/dynamic/loop_child.c number of iterations is 2000 (and not 1000) MPI_Comm_disconnect is invoked by both parent in loop_spawn.c : MPI_Comm_free(&comm_merged); MPI_Comm_disconnect(&comm_spawned); and children in loop_child.c : MPI_Comm_free(&merged); MPI_Comm_disconnect(&parent); is there any possibility you are running a different test called loop_spawn or an older version of the dynamic/loop_spawn test from the ibm test suite ? Cheers, Gilles
