1. Proposed title of this feature request: Scaling down configurations in Cluster autoscaler
2. What is the nature and description of the request? Ability to configure how cluster scales down the nodes is required in Hypershift API for the guest clusters.
3. Why does the customer need this? (List the business requirements here)

When there's a certain period of low traffic, the pods are removed and CAS starts removing the nodes. In the situations where intermittent traffic reduction or planned scale up to support blast of traffic, customers need to keep the nodes instead of scaling down and then again scaling up. 

Customers used stand-alone OCP have configured this https://docs.openshift.com/container-platform/4.15/rest_api/autoscale_apis/clusterautoscaler-autoscaling-openshift-io-v1.html#spec-scaledown. 

HyperShift clusters are OpenShift clusters but due to lack of these configurations, customers are unable to use Hypershift / HCP.

4. List any affected packages or components.

Cluster Autoscaler, HCP.