shit idea 💡 🤔 😕 😳 😐 😅 💡 time!!!
pbs/slurm cluster where all the desktops are nodes.
but a job can only run 2 hours max and nodes start draining at 6am. resume at 6pm.
for i in desktops ; do ssh $1 “top | head | grep blah | echo “can resume!” ; done
or jobs can run 12 hrs. u got all day to submit to queue..