The cable industry, by all indications, is doing a good job of transitioning to IPv6. Indeed, the new addressing scheme - which creates an almost limitless number of IP addresses - has systematically been implemented by MSOs, especially at the tier 1 level.These big operators - with
Comcast (NASDAQ:
CMCSA) leading the charge - are pursuing a "dual stack" approach in which legacy IPv4v and IPv6 run in tandem. The concept is simple: Keeping the two addressing schemes segregated avoids a world of complexity and allows the transition to move forward more quickly.However, there are a couple of questions with which the industry must deal. These questions are particularly pertinent to operators that due to size, manpower limitations and/or corporate agendas aren't as advanced in their IPv6 plans as others. Two of the key issues:
- Can IPv6 be implemented in a way that better sets up the operator for new and/or expanded business opportunities or allows its network to run more efficiently - or is it solely an address enhancement technique? In other words, should an operator's IPv6 program have goals beyond simply adding addresses?
- Should operators use an interim technology called Carrier Grade Network Address Translation (CGN) that allows IPv4 addresses to be extended well into the future?
Not Just IPv6These two very different questions speak to the complexities - and, perhaps, opportunities - of IPv6. Chip Popoviciu, the president and CEO of
Nephos6, suggests that service providers who have not yet gotten deeply enmeshed in IPv6 rollouts won't be able to implement the new addressing scheme in a manner that goes beyond the core function of extending the address pool.The idea is that it is possible to do things beyond extending the address pool - if the operator has the time. Presumably, that is still possible for operators with a store of IPv4 addresses and, thus, no pressing need. For instance, IPv6 address assignments could be coordinated with management of virtual local area networks (VLANs) or rollouts of software-defined networks (SDNs). The result could be a logically more cohesive and therefore more easily managed network.In some cases, it may be possible to garner these secondary advantages. "So, go for the right approach, do it right and make the most out of it," wrote Popoviciu in response to emailed questions. "That does not mean only leveraging the larger space, it can simply mean launching a long-discussed new billing model or a long overdue customer engagement model. IPv6 is a new beginning; don't make it legacy from day one."John Brzozowski, a fellow and chief architect for IPv6 at Comcast, agrees that IPv6 provides the opportunity for operators to grow.
"We are using IPv6 now not just as a tool to [increase addresses]," he said. "It’s strategic, leveraging as a foundation for innovation."Timing, of course, is key. Jeremy Schmeichel, a network engineer for
IBBS, concurs with the common wisdom that dual stack is the most elegant approach to the conversion to IPv6. He points out that an issue that could get in the way of smooth implementation - and, perhaps, the business initiatives that are made possible by the new addressing scheme - is the timing of upgrades to customer premises equipment (CPE). The challenges in this area, he said, could be more complex than on the commercial side.While change-outs of residential equipment are under control of the cable operator, enterprise customers may have other ideas. "Businesses may not be capable of changing their CPE, at least on the operators' schedule," he said. "Operators have to be flexible and mindful of where their [commercial] customers are."There also are CPE concerns on the residential side, said Craig Sprosts,
Nominum's general manager for fixed broadband solutions. "There is quite a bit of equipment that needs to be IPv6-compliant," he said. "The biggest hole in the tent typically is [that] the CPE have such a long replacement cycle. Operators may not swap cable modems for five or 10 years, who knows how long. Really, at this point it is largely about working with vendors and making sure every piece of equipment supports IPv6."
CGN: Another Complicating FactorAnother issue that is important for operators to think about is the use of CGN as a stopgap on the road to full IPv6 implementation. Network address translation (NAT) techniques enable the use of one IPv4 address to serve multiple end users. CGN - robust NAT with the bells and whistles carriers and enterprises need - can use the approach to stretch the potential time that IPv4 can survive. The downside is that NAT delays IPv6 and the advantages it offers.Gina Nienaber,
Cisco's (NASDAQ:
CSCO) marketing manager for service provider routing and switching, IPv6, said that in addition to CGN's limitations, it is an expensive solution because of the memory necessary to sort out the multiple end points served by each IPv4 address. "It has to remember more states," she said. "That requires more memory, more blades, and costs the service providers more money."Brzozowski agrees that CGN is not an optimal solution. "At this stage of the game, anyone who has gone down the path of evaluating or deploying [CGN] have gotten a surprise and found that it is not all it's cracked up to be," he said. "Real life experience puts a whole new light on the importance and timing of IPv6."The cable industry for the most part has settled on the dual stack approach. That doesn't mean that no issues remain. There is some room for alternate solutions - and the possibility that savvy rollout of IPv6 can be parlayed to provide other benefits. The clock is ticking at different speeds for each operator.
Carl Weinschenk is the Senior Editor of Broadband Technology Report. Contact him at [email protected].
