Siegmar, this is a known issue that is tracked at https://github.com/open-mpi/ompi/issues/1998
Cheers, Gilles On Sunday, August 28, 2016, Siegmar Gross < siegmar.gr...@informatik.hs-fulda.de> wrote: > Hi, > > I have installed openmpi-dev-4691-g277c319 on my "SUSE Linux > Enterprise Server 12 (x86_64)" with Sun C 5.14 beta and gcc-6.1.0. > Unfortunately I get an internal error for all my spawn programs. > > > loki spawn 147 ompi_info | grep -e "Open MPI repo revision" -e "C compiler > absolute" > Open MPI repo revision: dev-4691-g277c319 > C compiler absolute: /opt/solstudio12.5b/bin/cc > loki spawn 148 > > > loki spawn 151 mpiexec -np 1 --host loki --slot-list 0:0-5,1:0-5 > spawn_master > > Parent process 0 running on loki > I create 4 slave processes > > [loki:10461] [[46948,1],0] ORTE_ERROR_LOG: Unreachable in file > ../../openmpi-dev-4691-g277c319/ompi/dpm/dpm.c at line 426 > -------------------------------------------------------------------------- > At least one pair of MPI processes are unable to reach each other for > MPI communications. This means that no Open MPI device has indicated > that it can be used to communicate between these processes. This is > an error; Open MPI requires that all MPI processes be able to reach > each other. This error can sometimes be the result of forgetting to > specify the "self" BTL. > > Process 1 ([[46948,1],0]) is on host: loki > Process 2 ([[46948,2],0]) is on host: loki > BTLs attempted: self tcp > > Your MPI job is now going to abort; sorry. > -------------------------------------------------------------------------- > [loki:10461] *** An error occurred in MPI_Comm_spawn > [loki:10461] *** reported by process [3076784129,0] > [loki:10461] *** on communicator MPI_COMM_WORLD > [loki:10461] *** MPI_ERR_INTERN: internal error > [loki:10461] *** MPI_ERRORS_ARE_FATAL (processes in this communicator will > now abort, > [loki:10461] *** and potentially your MPI job) > > > > > loki spawn 152 mpiexec -np 1 --host loki --slot-list 0:0-5,1:0-5 > spawn_multiple_master > > Parent process 0 running on loki > I create 3 slave processes. > > [loki:10482] [[46929,1],0] ORTE_ERROR_LOG: Unreachable in file > ../../openmpi-dev-4691-g277c319/ompi/dpm/dpm.c at line 426 > -------------------------------------------------------------------------- > At least one pair of MPI processes are unable to reach each other for > MPI communications. This means that no Open MPI device has indicated > that it can be used to communicate between these processes. This is > an error; Open MPI requires that all MPI processes be able to reach > each other. This error can sometimes be the result of forgetting to > specify the "self" BTL. > > Process 1 ([[46929,1],0]) is on host: loki > Process 2 ([[46929,2],0]) is on host: loki > BTLs attempted: self tcp > > Your MPI job is now going to abort; sorry. > -------------------------------------------------------------------------- > [loki:10482] *** An error occurred in MPI_Comm_spawn_multiple > [loki:10482] *** reported by process [3075538945,0] > [loki:10482] *** on communicator MPI_COMM_WORLD > [loki:10482] *** MPI_ERR_INTERN: internal error > [loki:10482] *** MPI_ERRORS_ARE_FATAL (processes in this communicator will > now abort, > [loki:10482] *** and potentially your MPI job) > > > > > loki spawn 153 mpiexec -np 1 --host loki --slot-list 0:0-5,1:0-5 > spawn_intra_comm > Parent process 0: I create 2 slave processes > [loki:10500] [[46915,1],0] ORTE_ERROR_LOG: Unreachable in file > ../../openmpi-dev-4691-g277c319/ompi/dpm/dpm.c at line 426 > -------------------------------------------------------------------------- > At least one pair of MPI processes are unable to reach each other for > MPI communications. This means that no Open MPI device has indicated > that it can be used to communicate between these processes. This is > an error; Open MPI requires that all MPI processes be able to reach > each other. This error can sometimes be the result of forgetting to > specify the "self" BTL. > > Process 1 ([[46915,1],0]) is on host: loki > Process 2 ([[46915,2],0]) is on host: loki > BTLs attempted: self tcp > > Your MPI job is now going to abort; sorry. > -------------------------------------------------------------------------- > [loki:10500] *** An error occurred in MPI_Comm_spawn > [loki:10500] *** reported by process [3074621441,0] > [loki:10500] *** on communicator MPI_COMM_WORLD > [loki:10500] *** MPI_ERR_INTERN: internal error > [loki:10500] *** MPI_ERRORS_ARE_FATAL (processes in this communicator will > now abort, > [loki:10500] *** and potentially your MPI job) > loki spawn 154 > > > I would be grateful, if somebody can fix the problem. Thank you > very much for any help in advance. > > > Kind regards > > Siegmar > _______________________________________________ > users mailing list > users@lists.open-mpi.org > https://rfd.newmexicoconsortium.org/mailman/listinfo/users >
_______________________________________________ users mailing list users@lists.open-mpi.org https://rfd.newmexicoconsortium.org/mailman/listinfo/users
