Here at SDN World Congress, just a year ago, ten global operators created the framework of NFV with their Call For Action paper, and a new white paper was released here that renews and refines that call. I had the opportunity to read the new paper and to talk with many of those who authored it. There’s still the same flavor of improved capex through hosting of functions on servers, but there’s a new emphasis on opex and further emphasis on innovation.
I’ve been involved in NFV, as one operator noted, “from almost the first” and one thing that particularly gratifies me about both my conversations and from the “Perspectives” paper is the increased focus—not just focus on capex or opex or even on innovation and new services, but new recognition of the value of open source and openness in general, and new understanding of the details, the specific needs.
There’s been a lot of progress in the year that’s passed. Many of the same operators who signed the Call for Action saw a demonstration of NFV that was both open and conformant to ETSI specifications. I hope that they were gratified by the realization, at least in prototype form, of their visions. I also hope they understand that their visions are driving the greatest revolution in networking since the Internet—in some ways an even greater one.
When the Internet burst out in the 1980s from an academic/scientific curiosity to the ultimate conduit for experience delivery, it was riding into a hole in the vast ecosystem of telecommunications. We’d made the transformation to the digital age but the majority of users did nothing that a tin can and string might not have supported. There was no consumer data application, no justification for the bits we could produce in increasing quantities. We had a pond full of guppies and no pike to eat them.
The pike are eating now, of course. OTTs exploited the best deal for delivery of content that ever was, and became so much the face of telecom that many users think they get the Internet from Google or Yahoo and don’t have a clue that their carrier does anything more than send them their mobile or TV bill. That’s the real challenge that the Internet created—an ecosystem wasn’t created as much as exploited, and the exploitation hasn’t been entirely healthy.
DT said that they were going to become a software-defined provider, and that theme is reflected in the new white paper too. Some of the motivation for that transition is the fact that automation is the only way to control the cost per bit as the retail price slips at 40 to 50% per year. Some is due to the expectation that software and servers will supplant increasingly expensive “middle-box” and service layer appliances. Some is because the operators realize they will have to participate in the services of the future as much as they carried and created the services of the past.
It’s easy to oversimplify the changes we now face. We can say that we’ll do this with SDN or that with NFV and toss in a little cloud. We have a trillion dollars in assets out there, millions of workers whose knowledge of business and network technology are now threatened. The guppies in the pond have to grow their own teeth and confront the pike, too. Can they compete where the money is, and could they sustain their critical place in our new ecosystem if they can’t step up and take charge of some of the prime real estate of our metaphorical pond?
Software-defined operators are agile, not ossified. It’s easy to be brave now, at the start of our massive changes, and declare that we’re all for them. Gandalf, in Fellowship of the Ring, has some advice; “Let him not vow to walk in darkness who has not seen nightfall.” We’re seeing operators facing the darkness of change, maybe mostly because it’s getting even darker where they are now, in a role that is commoditizing toward near-zero incremental bandwidth revenues while costs don’t fall nearly as fast.
There are perhaps twice the number of authors to this new NFV paper than were behind the old, and those I’ve talked with were determined to lead the way toward our new network. The challenge for them will be to address the problem that confounds all visionary leaders—the mob effect. For the first five minutes of the march to the barricades, the crowd hears the leaders’ message and those who find it resonates run along with those leaders. Another ten minutes and the crowd can’t hear those leaders any longer, they’re just the mob. And we have millions, no billions, who have to follow somehow.
The first NFV ISG documents are now available, but if you read the new paper you know that while they’re necessary conditions for our facing our industry’s future, they’re not sufficient conditions. Development, software architectures, business planners, and even business partners will now have to jockey around to create and fill new eco-niches. We have to build a new set of technology relationships and a new set of business relationships within the operators and without, and it’s going to take time.
Most of all, though, it’s going to take guts. I’ve been in this industry now for almost 50 years, from the days of analog modems, punch cards, and half-duplex lines. I know what it’s like to see your world changed, your skills obsoleted. If you want to be on the leading edge you have to reinvent yourself on the average every two years. It’s never fun, but that’s what we now have to do if we’re going to take our place in the pond. Otherwise, we’re pond scum and not the “higher layer” we hear about.