On Thu, 2017-01-05 at 12:59 +0200, Graham Inggs wrote: > On 05/01/2017 12:05, Santiago Vila wrote: > > Status: FAILED > > Output: > > ------------------------------------------------------ > > -------------------- > > There are not enough slots available in the system to > > satisfy the 2 slots > > that were requested by the application: > > test-yaspgrid-yaspfactory-1d > > > > Either request fewer slots for your application, or make > > more slots available > > for use. > > ------------------------------------------------------ > > -------------------- > > I started seeing similar errors in other MPI applications since the > upload of openmpi 2.0.2~ to unstable. > > The solution was to add --oversubscribe to the mpirun command line,
Yes, it looks like OpenMPI changed the default :-/ > The bug should be reproducible with sbuild on a single CPU virtual > > machine. > > It always fail for me (I tried 10 times in different autobuilders). > > If I understand correctly, --oversubscribe should be needed in your > case where you have fewer CPUs than the number of processes > requested, but I was seeing the errors even when there were more than > enough CPUs available. On my laptop with 4 cores + HT (so 8 threads), I see `mpirun` complain once I start more than 4 processes, i.e. more processes than real cores. Did you count threads or cores when you tried? Ansgar -- debian-science-maintainers mailing list debian-science-maintainers@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/debian-science-maintainers