[OMPI users] Intermittent failure when launch application linked with OpenMPI 3.1.1

2018-10-04 Thread David Whitaker
Hi, When launching an application linked with OpenMPI 3.1.1 using the line: srun --mpi=pmi2 --distribution=arbitrary --cpu_bind=map_cpu:0,2,4,6,8,10,12,14,16,18,20,22,24,26,28,30,32,34,36,38,40,42,44,46,48,50,52,54,56,58,60,62,64,66,68,70,72,74,76,78,80,82,84,86,88,90,92,94,96,98,100,102,104

[OMPI users] Memory Leak in 3.1.2 + UCX

2018-10-04 Thread Charles A Taylor
We are seeing a gaping memory leak when running OpenMPI 3.1.x (or 2.1.2, for that matter) built with UCX support. The leak shows up whether the “ucx” PML is specified for the run or not. The applications in question are arepo and gizmo but it I have no reason to believe that others are not af

Re: [OMPI users] Cannot run MPI code on multiple cores with PBS

2018-10-04 Thread Jeff Squyres (jsquyres) via users
Note that what Gilles said is correct: it's not just the dependent libraries of libmpi.so (and friends) that matter -- it's also the dependent libraries of all of Open MPI's plugins that matter. You can run "ldd *.so" in the lib directory where you installed Open MPI, but you'll also need to "l

Re: [OMPI users] Cannot run MPI code on multiple cores with PBS

2018-10-04 Thread John Hearns via users
Michele, the command is ldd ./code.io I just Googled - ldd means List dynamic Dependencies To find out the PBS batch system type - that is a good question! Try this: qstat --version On Thu, 4 Oct 2018 at 10:12, Castellana Michele wrote: > > Dear John, > Thank you for your reply. I have

Re: [OMPI users] Cannot run MPI code on multiple cores with PBS

2018-10-04 Thread Castellana Michele
Dear John, Thank you for your reply. I have tried ldd mpirun ./code.o but I get an error message, I do not know what is the proper syntax to use ldd command. Here is the information about the Linux version $ cat /etc/os-release NAME="CentOS Linux" VERSION="7 (Core)" ID="centos" ID_LIKE="rhel f

Re: [OMPI users] Cannot run MPI code on multiple cores with PBS

2018-10-04 Thread Gilles Gouaillardet
In this case, some Open MPI plugins are missing some third party libraries, so you would have to ldd all the plugins (e.g. the .so files) located in /lib/openmpi in order to evidence any issue. Cheers, Gilles On Thu, Oct 4, 2018 at 4:34 PM John Hearns via users wrote: > > Michele one tip: lo

Re: [OMPI users] Cannot run MPI code on multiple cores with PBS

2018-10-04 Thread John Hearns via users
Michele one tip: log into a compute node using ssh and as your own username. If you use the Modules envirnonment then load the modules you use in the job script then use the ldd utility to check if you can load all the libraries in the code.io executable Actually you are better to submit a sh