Inter-core request forwarding for voltha

To support high availability it is essential that tasks be evenly distributed among all cores. Each voltha core is responsible for managing multiple logical switches. As GRPC requests are received by the proxy in all cases except the addition of a new logical device, the request could fall on any core. A forwarding capability is required that can decode the GRPC payload and make the appropriate forwarding decision to delegate the work to the core handling the device.

Acceptance Criteria:

• With multiple voltha instances running, any request from a NBI API should be served successfully.