The initial deployment of the three-pod design consists of three clusters, with one cluster in each of the pods. These clusters scale up as needed by adding ESXi hosts, while the pods scale up by adding additional clusters. The separation of management, edge, and resource functions into individually scalable pods allows CSPs to plan capacity based on the needs of the specific function hosted by each pod, providing greater operational flexibility.
For best practices, initial deployment requires a minimum of four hosts per cluster, for a total of twelve hosts. This sizing recommendation provides balance between the implementation footprint and resiliency, while maintaining the operational requirements necessary for each of the pods.
The resource and edge clusters are sized in accordance with the VNFs and their respective networking requirements. CSPs must work with the VNF vendors to gather requirements for the VNF service to be deployed. This information is typically available in deployment guides and sizing guideline documents
As more tenants are provisioned, CSPs must make additional resources available to the Resource pod to support VNF growth. Tenant administrators manage the allocation of OvDC resources to their VNFs to take advantage of the added resources. The increase in VNF workloads in the Resource pod can imply an increase in North-South network traffic, which in turn requires scaling up the ESG in the Edge pod by adding compute resources. CSPs must closely monitor and manage the resource consumption and capacity availability of the Edge pod as the ESGs are scaled.