Current documented guidelines for CPU, Memory and Disk Storage is based on a single all-or-nothing deployment of ONAP. This is no longer useful or desirable as many consumers of ONAP deploy a subset of the components and use the deployment for specific use usecases. It has been discussed, and agreed upon a number of times, that at minimum we can provide 2-3 different sizings of ONAP.

Examples would be:

• small for development and functional testing of core components
• medium for HA, resilience and some throughput testing of core components
• large for HA and resilience and throughput testing of all components