In the past, we’ve spent some time explaining what we believe is the first great use case for carrier SDN. Well, let’s continue that discussion with the first great use case for NFV. Driven by global customer interest and the multi-domain orchestration capability of Cyan’s Blue Planet, we’re proud to introduce the concept of NFV-enabled Ethernet. And we believe it’s critical to the future growth of service provider revenue streams on existing network services.
Let us explain.
Common off the shelf (COTS) servers are comprised of commodity CPUs that have become ubiquitous. Not only are they deployed in mega-scale data centers but also in many PoPs and COs, as operators convert hundreds or thousands of such premises to generic telco data centers.
A centralized data center provides economies of scale, but distributed data centers provide resiliency, agility and better performance for both latency or performance sensitive edge applications, as they tend to be closer to the customer. Now imagine, evolving or even replacing Customer Premise Equipment (CPE) with generic servers that are virtualized or adding servers to NIDs. Every customer location turns into a compute resource connected through a simple interface.
This concept, called distributed-NFV or more commonly, virtualized CPE, enables operators to shift many network functions from gateways or appliances toward the carrier’s own network, by running these virtual network functions (VNFs) on COTS servers. With such an approach network functions such IP routing, firewalls, encryption, IP address management (DHCP), address translation (NAT), firewall, and various set top box functionalities (live TV, VoD, etc.), etc. can be easily maintained and upgraded from the most efficient location.
How do we turn this concept into reality? Well, there is a lot going “under the hood”, but Cyan is demonstrating this technology, along with RAD, Fortinet, and Certes at the Light Reading Big Telecom Event this week. In this use case, which is ETSI approved and sponsored by CenturyLink, we show how Cyan’s Blue Planet can orchestrate firewall and encryption VNFs in a distributed architecture (RAD edge device).
But the real power of this demonstration goes beyond just the instantiation of VNFs at the edge. Let’s look at the RAD edge device. It serves as both an Ethernet NID and also contains a generic compute server within it. So now, Cyan customers can use Blue Planet to both set-up an Ethernet service automatically from end-to-end, but also provision new virtualized services on top of that Ethernet circuit. This is a concept we call NFV-enabled Ethernet made possible because Blue Planet is both a WAN and NFV orchestration system. No NFV orchestrator can do this on its own.
Today, many customers already use Blue Planet to provision carrier grade E-Line and E-Lan services, but now, Blue Planet can enrich Ethernet service delivery by also orchestrating and managing the following:
- Downloading and installing OpenStack on bare metal, server based CPEs
- Creating virtual networks on the CPEs
- Downloading and onboarding multiple VNFs
- Service chaining these VNFs
- Running virtualized EMSs for each VNF in a central location
- Tracking the health, utilization and performance metrics of all virtual resources
- Dynamically adjusting virtual and physical resource allocation
- And many, many more all under a single pane of glass
With Blue Planet, operators can quickly move up the chain and secure new revenue streams by offering enriched NFV-enabled Ethernet services. Additionally, the service provider controls the user experience by provisioning both the carrier grade network as well as the over the top software applications (VNFs) that use-to-be enterprise hardware appliances.
I will personally be at the Big Telecom Event. Make sure you drop by and request a demo.