This month saw a major step in the move toward IPv6 as over 3,000 websites including heavy hitters like Google, YouTube, Facebook and Yahoo began supporting IPv6 on a permanent basis. On the day of the World IPv6 Launch, Arbor Networks reported a 20% jump in IPv6 across the 15 service provider networks it monitors. However, even with that 20% bump, IPv6 only amounted to 0.1% of all IP traffic. Which gives you an idea of just how far we have to go with IPv6 adoption.
Indeed, that point was hammered home repeatedly at a World IPv6 Launch event organized by the Internet Society Hong Kong. APNIC chief scientist Geoff Huston made the point repeatedly that despite the fact that IPv4 is not even an option in Asia Pacific for ISPs seeking new addresses, IPv6 adoption is by no means a given, and there's still plenty of room to get it wrong.
Bearing the cost
Huston singled out last mile providers (both wireline and wireless) as "abysmal failures" in terms of IPv6 readiness primarily because of the lack of economic incentive. Put simply, fixed-line margins are too low, the cost of running both IPv6 and IPv4 in 3.5G networks is too high, and they can't pass on those costs to consumers.
Complicating things is the fact that many ISPs now allow customers to buy their own modems, which raises a dilemma, Huston says: "How do you get them to upgrade when the CPE works just fine and will last ten years? How do you explain to them they have to upgrade to IPv6?"
It's an interesting dilemma, not least because it puts ISPs in the unenviable position of having to explain technology to customers who don't want to hear about technology. Just as customers don't care what the difference is between 3G and 4G (apart from the cost and the data cap), they're probably less likely to care about the difference between IPv4 and IPv6.
ISOC Hong Kong is concerned enough about the education issue that it's embarked on a market education campaign in conjunction with the SAR government to explain what IPv6 is, why it's important and how to enable it. What the education materials (which include a cartoon guide and a website) don't mention what it might cost to upgrade it, or whose responsibility that cost is. The implication seems to be that if the CPE has to be replaced, either the consumer or ISP will have to cover that cost.
On the bright side, Ovum principal analyst David Krozier says he's been assured by most vendors that IPv6 is mainly a software or firmware issue for most CPE, apart from older home residential gateways. So that may be a minor issue.
But not so minor that it can be ignored or left to the last minute. If customer-centricity really is the order of the day, ISPs of all stripes had best be prepared to not only enable IPv6 in the last mile, but also keep their customers in the loop about why IPv6 matters, and what they stand to gain.