The following design principles communicate a summary of the details that form the basis for the vCloud NFV Edge reference architecture using vCloud NFV components.
Deployment Flexibility
Starting with a low management overhead at the Edge sites, the architecture provides flexibility for including a two vSphere cluster configuration (separate compute cluster and Edge cluster) at Edge sites requiring higher performance with NSX Edge Nodes. In cases where the number of compute nodes at the Edge site are limited, it possible to have a combined compute and Edge cluster where the NSX Edge Nodes run as VMs on the same compute cluster as the workloads
Multi Tenancy and Advanced Networking
Within an Edge site, it is possible to house one or more tenants managed from the regional or core data center. This is realized using the multi-tenant capabilities of NSX-T (T1 and T0 routers) and the integration with VMware Integrated OpenStack (VIO).
Workload Acceleration
To run user plane functions requiring lower latency and higher throughput at the Edge sites, the N-VDS (Enhanced) virtual switch can be configured on the hosts where these functions are to run. The UPF VMs can then be configured to run on top of the N-VDS (Enhanced) switch.
Real-time Integrated Operational Intelligence
By using a framework that continuously collects data from local and distributed agents, vCloud NFV Edge Reference Architecture provides the capability to correlate, analyze, and enable day 2 operations. This analytics engine can be used with existing assurance engines for closed-loop. In addition, this analytics engine can be deployed in a distributed fashion in near-Edge sites towards smarter WAN bandwidth management and real-time closed-loop assurance.
Orchestration Integration
vCloud NFV OpenStack Edition provides the flexibility to integrate the distributed management components, such as VIMs, with open-source-based Orchestrators such as ONAP, OSM, and third-party Orchestrators using industry standard interfaces.
