1. Provide Descheduler profile based on utilization of nodes
2. The current Descheduler operator does not deschedule based on the actual utilization of the nodes. We would like to see a profile based on cpu and memory of a node. If possible and reasonable, also a profile based on network traffic of a node.
3. Customer is an Openshift Virtualization user looking for a DRS-like functionality. Currently, balancing out usage on the nodes is an exercise left to the customer.
4. Descheduler Operator