Customers may hit CPU limits before they reach limits imposed by the cloud provider. E.g Azure has a 250 pods\/node limit but each dash app requests 0.1 CPUs so DE would need to be running on nodes with > 25 CPUs before it's possible to hit that limit. Brining visibility into CPU usage alongside memory (existing) allows customers to plan infrastructure better. https://github.com/plotly/dekn/issues/6512
