Jeffrey Tunison wrote: > Is there a straightforward way to create a batch job that runs once on every node in the cluster? A wrapper around reboot configured as RebootProgram in slurm.conf?