FRONTEND - fix issue where if a compute node was in a down state and

an admin updates the node to idle/resume the compute nodes will go
instantly to idle instead of idle* which means no response.