You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
reproduces on orion with 16 nodes (but not with 4), so maybe scale/race issue?
Original command is module load hpcx-gcc && salloc -N 16 -p orion mpirun --display-map --bind-to core -mca pml ucx /usr/mpi/gcc/openmpi-1.10.3rc4/tests/osu-micro-benchmarks-5.2/osu_allgather, and the visible error is:
reproduces on orion with 16 nodes (but not with 4), so maybe scale/race issue?
Original command is
module load hpcx-gcc && salloc -N 16 -p orion mpirun --display-map --bind-to core -mca pml ucx /usr/mpi/gcc/openmpi-1.10.3rc4/tests/osu-micro-benchmarks-5.2/osu_allgather
, and the visible error is:Ran with ucx master, built with debug, and got the following:
The text was updated successfully, but these errors were encountered: