NFV is a great technology, but it is important to make sure that it gets its nap. There also is the chance that it will have a tantrum in the dairy aisle, according to Ray Watson, vice president for global technology at Masergy.
Of course, Watson didn't actually say that. Watson, who is delivering a Wednesday keynote at Light Reading's Light Reading's NFV and Carrier SDN event in Denver, used an extended toddler analogy to describe the current state of NFV. His keynote -- "Surviving the Terrible Toddler Years: The Good, The Bad, and The Ugly" -- suggests that NFV shares attributes with these sometimes adorable youngsters. They both enjoy periods of endearing preciousness. They both have unpredictable periods of troublesome behavior.
Above all, they both have awe-inspiring potential.
NFV will gain refinement as time goes on. Still ahead are finalization of business models and completion of standards. The most important ongoing challenge, however, is finding the right niches.
"Use cases are still being tested," Watson said. "We are still waiting for the killer use case. In the real world, not just the lab."
Right now, the key is for providers to listen to customers.
"The most basic question has to revolve around what business objectives are trying to be met," he told Telco Transformation. "It's not just technology for the sake of technology."
The managed service provider comes to virtualization in general and NFV in particular from a unique perspective. Last summer Watson told Light Reading founder Steve Saunders that Masergy wasn't burdened by the legacy infrastructure and full service responsibilities of a major carrier. This enables it to focus on a narrower range of functions and to develop platforms and products with the speed that web companies have brought to telecom. Only 71 days elapsed between the initial concept to rollout of its first product, Watson told Saunders. (See Masergy's NFV Journey.)
Masergy Communications Inc. has offered on-demand bandwidth since 2001 and customer self-provisioning and NFV since 2015. It currently offers router, firewall, WAN acceleration, SBC and managed encryption VNFs, with 27 more in its labs. Significantly, the initial goal of the NFV initiative was not direct cost savings. The initial drivers were agility, reduced shipping, easier coordination and less power consumption. Watson told Saunders that a surprising source of strength of NFV is among companies that generally have physical devices held up at customs.
Cutting costs was not on the initial list of drivers. It is compelling, of course. Watson said that regional multi-million-dollar companies are entering the market. That's a change from the multi-billion multinationals that formerly dominated advanced networks.
During his Denver keynote, Watson will describe three versions of NFV. One terminates at the carrier POP and a second at the virtual CPE (vCPE) at the customer premises. The third is the use of NFVs as an element of an SD-WAN. He will also talk about the benefits and shortcomings of each approach.
This is fluid technology, and an organization with varying needs -- say a large headquarters campus and small satellite offices -- can combine the approaches.
"All three are moving forward in parallel because the use cases for each are a bit different," according to Watson. "However, SD-WAN can actually be run on top of -- or in combination with -- either carrier- or prem-based NFV."
Once toddlers begin maturing, change happens quickly. That's another similarity with NFV.
"Mass adoption of virtualization and programmability will be more revolutionary to our industry than MPLS's arrival in the 1990s," Watson said. "We need to be ready. This means understanding how to protect our customers' assets. It means being able to report in real time on what their applications are doing. And it means moving at DevOps speeds."
— Carl Weinschenk, Contributing Writer, Telco Transformation