Ability to limit a service to a max number of CPUs or amount of RAM used.