I occasionally run into problems with space during a run on my cluster, which causes the job to fail and the node to be marked as unavailabl...