Provide a way to trigger restarts #194
Description
Restarting slurm daemons restarts the scheduler loop which it is nice to avoid, but it would also be good to provide a way for "external" functionality (e.g. stackhpc/ansible-slurm-appliance#737) to tell the role it needs to restart things for new config. Maybe we could replace the handlers with a restart playbook, and trigger that by setting a variable openhpc_restart_slurm or something? A conditional set_fact on an external task could be used to trigger that then.