in theforeman/foreman we have a formula how many puma workers we want to deploy, based on the CPU/ram of the machine