“There are things known and there are things unknown, and in between are the doors of perception,” wrote Aldous Huxley. That could be a description of the evolving tension between the perceptions of hype and reality of the network functions virtualization (NFV) market as it enters its important phase of commercialization.
The perception of some technology vendors and investors might have you thinking NFV is the next trillion-dollar market, poised to rescue cloud providers and network operators from the evils of proprietary hardware. Another perception from some of the most jaded analysts might be that it’s all marketing hocus pocus — corrupt slideware that has no basis in reality.
The reality, of course, is somewhere in between. The evolution of technology markets is messy, forever striking a balance between commercial interests, standards, and collaborative approaches such as open source. The growth path of NFV and its its cousin, SDN have their challenges, but our research indicates that the path has already been set and turning back is unlikely at this point.
We’ll be releasing some in-depth research in this area this week, in the first of our two-part Mega NFV report series. Let me share some of the survey research we conducted to gather a picture of where NFV is today. The SDxCentral Research team surveyed our readers about various NFV topics. There were 79 end-user respondents, including service providers, (46%), cloud service providers (14%), enterprise end users (14%) and a variety of other (24%) user communities.
Here’s what they said:
NFV Adoption Drivers
When asked to name the primary driver for NFV technology, the majority of respondents (50%) said it was NFV’s ability to “Deliver Agility and Flexibility.” “Accelerate Time-to-Market” was identified by 14%, while the remainder of the respondents chose savings as the primary reason to adopt NFV – 23% focused on the operational savings of NFV, while 13% said it was NFV’s ability to reduce capital expenditures that was so attractive.
So what kinds of functions would benefit from being virtualized the most? According to user respondents to the SDxCentral survey, L4-7 Services, such as firewalling and intrusion detection and intrusion prevention (IDS/IPS), are the best candidates, receiving 47% of the vote (respondents could pick two). Application suites, such as virtual evolved packet core (vEPC), virtual customer premises equipment (vCPE) and IP multimedia subsystems (IMS) received the second most votes, at 39%, followed by L2-3 Services, such as DHCP, DNS, routing, e.t.c., at 38%.
Service-based offerings, which have been described by the industry as good candidates for NFV (think Virtual Network Function as a Service (VNFaaS) and Virtual Network Platform as a Service (VNPaaS)) were chosen by less than a quarter (23%) of respondents.
Leading NFV Use Cases
When probed on the use cases (they could pick two) that were most appealing for NFV, an overwhelming 78% of respondents cited the software-defined wide area network (SD-WAN) or vCPE. 51% identified the virtualized mobile core (vEPC, vIMS, etc.), while 41% noted service-chaining as part of an SGi/Gi-LAN deployment as a good fit.
Seventy-eight of respondents to the SDxCentral survey feel NFV technology is starting to mature, but still needs more work; 18% feel it is “Not Mature at All and Needs a Lot of Work”. When probed further, respondents noted there were a variety of factors, which point to the early stage of the market, that are impeding the faster rollout of NFV.
Fifty-three percent of respondents said the difficulty certifying and integrating multi-vendor solutions was a barrier, which speaks to the immaturity of the overall NFV ecosystem. Twenty-nine percent of respondents said a lack of turn-key NFV integrated systems and general education on NFV architectures and best practices were hindering deployments, which also indicates how early the market still is. Interestingly, less than 10% of respondents (9% to be exact) felt they couldn’t see a compelling advantage for NFV over existing solutions.
Vendors: Take note. These response indicate an audience that believes in the tools of NFV, but doesn’t yet think the market is mature. They’re looking for the expertise and guiance of vendors and integrators to put the technology pieces in place.
The Role of ETSI
The European Telecommunications Standards Institute (ETSI) hosts the NFV Industry Specification Group (ISG), the braintrust behind the original NFV architecture. Since its founding in 2012, the Group’s membership has grown to more than 270 companies, including 38 of the world’s major service providers. Together, the working group has published standard terminology and reference use cases designed to improve the performance, efficiency and agility of NFV deployments. When asked if the ETSI ISG architecture was seen as the primary source of standards for the NFV ecosystem, respondents overwhelming replied “Yes” (91%).
The primary framework published by the ISG Working Group is NFV management and orchestration (MANO); it is designed to enable the flexible on-boarding and ongoing management of new network services. It has been several years since its release, and its relevancy has started to be called into question. Most respondents (82%) felt NFV MANO was “Coming to Maturity, But Still Needed More Work”. 18% felt the framework still “Needs a Lot More Work”.
So where is NFV based on some of these perceptions? Our fact gathering leads me to believe that NFV has exited the speculative phase and has entered its most painful phase: commercial deployment. Going into the research, I expected more skepticism and challenges. But our fact finding has revealed enthusiasm from the end users. NFV and ETSI MANO, in particular — appear to have garnered an accepting and believing audience.