In a distributed cloud topology, VNFs are stitched across the different cloud islands to delivery services. Each site can also have multiple instances of the NFV environment for redundancy and scale.
Each site is a standalone island with its own instantiation of the Management, Resource, and Edge Pods. Multi-site designs are typically used for load distribution and for high availability in case of catastrophic site failures.
The design assumes an active-standby configuration for geographic redundancy. The geographic redundancy design can be implemented with proximal sites or with larger physical distances, as network latency between sites must be considered.