[torqueusers] torque+blcr+openmpi

Anton Starikov ant.starikov at gmail.com
Tue Feb 23 03:15:27 MST 2010


Can anyone provide example of checkpoint script for torque which deals with open-mpi checkpointing?

I have little doubts how to checkpoint parallel jobs, because they can't to be checkpointed as whole tree with pbs_mom as parent by cr_checkpoint, but should be chekpointed separately by ompi-checkpoint.



More information about the torqueusers mailing list