What's Next with IPv6?

IPv4 Depletion is Just the Start

We recently held the annual executive meeting of the Rocky Mountain IPv6 Task Force – which primarily consisted of eating sandwiches around Scott Hogg’s kitchen table – and a topic on the agenda was a discussion of future directions for the Task Force. An opinion was expressed that the RMv6TF’s mission would be completed in a couple of years, and perhaps we should begin making preliminary plans for disbanding. IPv4 depletion is a reality, so the only message left for us is “We told you so.” Almost all service providers have IPv6 deployment plans underway and at some stage of completion. Any IP networking class worthy of your money teaches IPv4 and IPv6 as equally important versions of IP. Is there anything of importance left for an IPv6 advocacy group to do?

Yes! There is still much to do, and industry activity related to IPv6 has only begun. Broadband service providers certainly understand the business drivers for IPv6 and are (for the most part) aggressively planning for it. The same goes for carriers and upper-tier ISPs. IPv6 has long been available on most operating systems and major routers, although feature parity with IPv4 is atypical. Vendors of peripheral support systems (security and management and the such) are catching up – increasingly, a lack of IPv6 support and lack of an IPv6 roadmap is an automatic rejection on customer purchase requirements.

So what’s next?

Content providers and content delivery networks are the next big area for IPv6 deployment. There are a few big names (and some not so big names) who have prominent IPv6 efforts, such as Google, YouTube, Netflix, and Limelight, but for the most part content is available across the Internet only over IPv4. Most of these entities know that they will need to support a growing body of IPv6 users in the near future, but they are either waiting for compelling IPv6 user growth or are wary of what IPv6 connectivity weirdness might do to their business. Or both.

100% of my IPv6 consulting clients right now are service providers, but I’m beginning to get calls from content providers looking for help in getting a plan together. My purely casual view is that this segment of the networking industry is where most service providers were about five years ago.

Somewhere behind content providers and CDNs are enterprises. Again there have been a few prominent IPv6 deployments in enterprises, such as Bechtel; British banker HSBC has also been making news with their IPv6 plans, due to their large Asian footprint. But these are a small minority; very few enterprises see a business case for deploying IPv6 in their networks. They’re beginning to ask questions, but that’s about it.

When I hear from enterprises at all, it is not even as formal inquiries but just causal questions: “Do you think I will ever have a reason to install IPv6? Is there any business case at all?” These are the kinds of questions service providers were asking me about 10 years ago.

There is indeed a business case for both content providers and enterprises, and it’s not what you might expect. I’ll write about that in a subsequent post. My point here, though, is that there will be a role for IPv6 advocacy groups like RMv6TF for many years to come as the audiences change from service providers to content providers to enterprises.

There is one last factor that will keep IPv6-related projects going for years to come: Decommissioning IPv4. Right now all of us operate under the safe assumption that IPv4 and IPv6 must coexist for the “foreseeable future” because none of us can say with any certainly how fast IPv6 usage will surpass IPv4. But the reality is that managing two versions of IP in a network is operationally expensive and will increase operational errors. Transitional technologies add capital expense to the network and negatively impact security and performance. As a result network operators are going to begin, in a few years, to push IPv4 out of their networks. Broadband providers will be watching usage data with an eye to what point at which it will be cheaper to lose their few IPv4 holdout customers than to continue supporting IPv6 at their customer edge.

The day when network operators begin shutting down IPv4 will come earlier than many expect. Yep, still lots to do.

Join the Network World communities on Facebook and LinkedIn to comment on topics that are top of mind.
Must read: 10 new UI features coming to Windows 10