NFV Group Preps Its Afterlife
The leadership team at the ETSI network functions virtualization (NFV) Industry Specifications Group (ISG) has decided that its original, self-imposed two-year timeline is too short and is now considering the best way to extend its life beyond 2014.
The core of the carrier-driven group first met formally in October 2012 in Germany, and the European Telecommunications Standards Institute (ETSI) group was born at the beginning of 2013 with a goal to complete its work (specifying NFV and identifying the business case) and disband by the end of 2014.
But NFV has become just about the hottest trend in the industry and the group's leadership team has realized that, one way or another, more needs to be done.
"A lot has been achieved in terms of work done and influence demonstrated," said Verizon Communications Inc. (NYSE: VZ)'s Prodip Sen, chair of the ETSI ISG, in an email to the group's near 200 members prior to a meeting in Malaga, Spain, last week. "It is clear to all that the vision we laid out is where the networking industry is collectively headed, though some may quibble about timelines and specifics... But given all the ongoing work, it is clear that our original goal -- of completing the work and disbanding in two years -- was a bit simplistic. The very strong industry attention which the NFV ISG has received, the large number of involved participants/players, and of course the identification of a lot of work that needs to be done, demands some form of continuation," he added.
So the leadership team has drafted a paper, titled "Proposal for NFV Phase 2," on a possible next step. Sen, though, made it clear that this was a starting point for discussion, and certainly not any kind of imposed decision. The aim is to discuss the future at the next meeting, to be held in Okinawa during May, and figure out what will happen come 2015.
The move makes sense: Despite a slow start, a great deal has been achieved during the past year and, as an excellent blog published by Ciena's Marc Cohn following the Malaga meeting shows, there are nine proof-of-concept projects underway and more use cases still to be tackled. (See Cohn's blog here.)
As many have noted, the ISG has made faster progress than some other telecom industry groups because it has been driven by the needs of network operators, which have pushed the vendor community to deliver to their specific requirements needs.
That doesn't mean that some don't want it to go faster. Dor Skuler, general manager of the CloudBand Business Unit at Alcatel-Lucent (NYSE: ALU), knows where he'd like the focus of NFV Phase 2.0 to be.
"We need to find a way to allow us to make decisions faster -- decisions that are not necessarily consensus-based," noted Skuler, who added that ETSI's rules require consensus decisions. "There is no consensus on some topics, but everyone involved is very passionate about making NFV happen and willing to collaborate to make NFV happen. And we need to stop worrying about standards but instead point towards open specifications such as TOSCA [Topology and Orchestration Specification for Cloud Applications]" to take things forward, added the AlcaLu man.
Skuler is not alone. Ciena's Cohn quotes AT&T Inc. (NYSE: T)'s Margaret Chiosi in his blog as saying: "NFV must overhaul the traditional standards adoption process, which typically requires years before implementations are available. We are seeking a Fast Fail Forward model that enables standards and their implementations to be iteratively developed, similar to the highly-successful open source software model."
Clearly, whatever the group decides to do will be critical to the ongoing development of the NFV ecosystem and industry watchers will be keen to see if Phase 2.0 remains carrier-led.
For more on NFV:
- Cisco CEO: It's Early Days for Virtualization
- TM Forum Tries ZOOMing to NFV
- Telefónica Unveils Aggressive NFV Plans
- Vendors Catch NFV Fever
- Alcatel-Lucent Lays Out Its NFV Plans
- Where Are We With NFV?
- Top 5 NFV Movers & Shakers
— Ray Le Maistre, Editor-in-Chief, Light Reading