Likely IPv6 Adoption Challenges Over the Next 2 Years
While there is strong evidence that cloud, service, content, residential and mobile providers in the U.S. have made the shift to IPv6 (see the HexaBuild IPv6 Adoption Report and infographic or the World IPv6 Launch infographic for more details) it is also clear that there is a huge gap in the adoption story. That gap is enterprise, commercial and small business adoption of IPv6. There are many reasons for this gap but I wanted to explore some of the most important ones and what needs to change over the next two years to close this gap.
Adoption among these business groups has been somewhere between slow and non-existent and I believe, for many, the idea of having to deal with IPv6 is still an incredibly low priority. For many IT organizations, IPv6 deployment may not even be on the list of things to do. For some, they feel they have enough public IPv4 address space for their company to run for many years (utilizing NAT44 at their Internet perimeter) and they see no reason to spend time or effort on learning IPv6. For others, they outsource their networking requirements and have no idea if they need IPv6 or not. Their managed service providers may not know how best to implement IPv6 even if their customers had requested it. Because their customers may not know or care about IPv6, the managed service provider can continue to ignore the issue. From their short-sighted perspective, it’s one less trouble ticket, one less project to complete, less operational expense and perhaps more profit for them (though this is arguably just delaying the inevitable expense of dealing with IPv6 to a later date).
The reasons behind the gap are easy to identify. Lack of knowledge (and the training that would mitigate it) of how to deploy IPv6 is first on the list (and, like the old cliché about real estate, perhaps even second and third as well). Lack of real industry experts to help the enterprise and Fortune 1,000 based companies complete a project successfully is a big problem too. Many enterprise, commercial and small businesses rely on resellers and system integrators to help them solve their technology problems and almost none of those services organizations are prepared to educate their customers on IPv6 much less design and deliver solutions that are IPv6 ready. This is among the most serious problems the IT industry faces (don’t take my word, listen to Vint Cerf on the Internet Society page) given that failure to embrace IPv6 will lead to entirely avoidable scalability and Internet connectivity and performance problems for enterprises going forward, and it seems few, if any, vendors or manufacturers currently see any upside to immediately addressing the problem. And this contemporary neglect is a significant distinction: Years ago, when many of the manufacturers of hardware and software added IPv6 support they spent time and effort on the new protocol perhaps expecting some sort of spike in sales because the technology was “new” or “industry standard” but then not much happened. They might have become somewhat disillusioned with how much IPv6 could or could not move the needle on helping them sell their product. There wasn’t any special new “thing” that IPv6 did outside of providing a crazy huge amount of public address space. Almost none of their customers were demanding IPv6 in the early days and the few that did were lower-margin business models with very high uptime and performance requirements (for instance government organizations like the Social Security Administration who first did their work in 2001). This didn’t provide the return on investment they desired but they knew they still needed support going forward so they shelved much of the innovation aspects of IPv6 and focused on minimum viable support. It left would-be IT adopters of IPv6 with a conflicting message from the vendors of “IPv6 is supported but don’t use it because it is not ready for prime time.” Really, what they were saying was their code and hardware wasn’t thoroughly tested with IPv6 in any significant way and you might now get the operational outcome and vendor support you wanted if you enabled IPv6. It was a tacit acknowledgement that their IPv6 work wasn’t really completed.
So, what has changed? Is there something different about 2018 as compared to years past? After all, it has been over 6 years since the World IPv6 Launch date where we turned on IPv6 and left it on. What has happened over the last 6 years to persuade a new generation of IT operators, admins and architects to finally look at IPv6 and say it makes the short list of things that need to get done soon?
I believe the following facts now impact the decision on if IPv6 gets implemented for many IT organizations:
- S. IPv6 adoption rate is just below 50% of overall traffic on the Internet.
- The three major US mobile providers are between 70% and 93% native IPv6 traffic on their networks.
- The majority of the content providers have enabled IPv6 (Google, Facebook, YouTube, Netflix, Yahoo, LinkedIn, etc.).
- All major cloud providers have IPv6 support available for their customers (though the comprehensiveness of their offerings vary).
- Residential providers have completed, or are close to completing, their IPv6 deployments in the U.S. (e.g., 67% of Comcast’s traffic is IPv6)
- The majority of U.S. residential customers have moved to DOCSIS 3.1 meaning those customers are capable of IPv6 enablement.
- The majority of significant networking manufacturers now have IPv6 support that is beyond the basics.
So, based on the above, what we have now a moment of critical mass in the case for enterprise IPv6 adoption: End-users are accessing Internet content on IPv6-only networks or preferring IPv6 when they do access content. Apple is requiring IPv6 support for apps to be approved in their App Store. This means developers of iOS apps need IPv6 to test and confirm their apps comply with that requirement. In addition, major business productivity apps (Microsoft Office 365 and Google GSuite) are both accessible over IPv6 meaning they prefer IPv6 for the majority of end devices. Other cloud services also have similar behaviors.
The gap is that many IT teams still have no idea how to log, audit, secure or even how to read an IPv6 address – not to mention the demands of running an operational data center or campus network that has IPv6 capabilities. Because all the services their end-users are accessing when they are away from their corporate office are happening over IPv6, getting a handle on IPv6 is becoming more important than ever for these teams. How do you deal with networking issues for remote employees accessing both the Internet and Internet-based work resources? Is your helpdesk able to debug and figure out what is going on? Can they even get access to the remote device if they don’t have IPv6 network connectivity themselves? More and more operational gaps start appearing with the lack of IPv6 in the enterprise, commercial and small business space. Lack of IPv6 training and knowledge makes the situation even worse. These companies can’t get help from their internal staff or from resellers and very few truly knowledgeable IPv6 consultants are available.
So, in the short term, it is likely that many companies will continue on with IPv4 as they struggle to learn IPv6 while the rest of the Internet moves on without them. It won’t be a pretty journey and we will likely see some major failures on the design and deployment side for these companies. So let’s not think for a moment that this transition isn’t important. Within the next two years (likely much sooner), IPv4 will be the minority of traffic on the public Internet in the U.S. Let that sink in for a moment. The network protocol you know best and currently rely on, support and maintain is NOT the protocol used by the majority of the Internet in the U.S. I think it is time we started talking about IPv6 education and training because our industry has fallen woefully behind in this area. We can cover that topic in another blog but get started with your journey to IPv6 today. After all, the rest of the Internet has already done so!
You can find me on twitter as @ehorley and remember…
IPv6 is the future and the future is now!