The move to IPv6 isn't an easy one, but those sticking with IPv4 are on board a sinking ship.

Owen DeLong

August 19, 2013

4 Min Read
Taking the Fear Out of IPv6

IPv6 can strike fear into the hearts of many in the Internet world, and not just because it sounds like a communicable disease.

IPv6 is still a new protocol. (Get tested here!) Deploying it in the average enterprise requires a substantial investment in training, education and planning. It also requires changes to established IT practices and often affects certain business processes and controls. Depending on equipment and software choices, there are often costs associated with hardware and software upgrades. In addition, homegrown applications will likely require updates.

Although this transition may seem daunting, it has been done before. Experienced system administrators remember the transitions through ARCNET, DECNET, Vines, AppleTalk, Novell NetWare and NetBIOS (a.k.a. Windows Networking) -- and finally to Internet Protocol.

Two things can be learned from these previous transitions: First, the IPv6 transition is not the first protocol switch, and it's not nearly as radical as some that preceded it. Second, this is not the first time two protocols have had to run on the same network at the same time.

Yes, the tunnel of IPv6 transition will be a bit scary, but once at the other end a world of unprecedented freedom and new capabilities await. It's a whole new path, but the sky is blue and you will hear the birds singing.

IPv4 approaches a dead end
To those not paying attention, continuing to run IPv4 seems pretty safe and comfortable. It's a lot like steaming along the dark northern Atlantic Ocean at high speed in an "unsinkable" ship. It feels really good and nothing dangerous can be seen. An early arrival in New York is expected.

But to those paying attention, it's clear that the IPv4 Internet has been on life support for about 20 years. Network Address Translation (NAT) address sharing degraded the Internet's end-to-end addressing without permanently solving the IPv4 address depletion issue. Multiple layers of NAT could kick the can further down the road, at the cost of further degrading and complicating systems (such as VoIP) that rely on end-to-end addressing. Furthermore, these degraded capabilities will cost significantly more to implement and maintain, so ISPs will likely increase costs to subscribers who cling to IPv4.

To make things even worse, we're reaching a point where adding new IPv4 customers is going to become virtually impossible, causing IPv4-reachable sites to become a continuously smaller fraction of the Internet. Today, almost everything on the Internet is reachable over IPv4. In a few years, both IPv4 and IPv6 will be required for universal access.

When I look at the real costs, trade-offs and problems created by ignoring IPv6 and trying to keep IPv4 going for another decade or so, it pegs the needle on my fear meter.

Lots of IT professionals will happily tell a different story. Some will say that an enterprise won't need to deploy IPv6 if it has already been allocated more IPv4 addresses than it needs. Some will say that there's no need to migrate to IPv6 because there aren't enough IPv6-capable users (or services) to matter. These points have an aura of truth: however, they're misleading and mostly irrelevant.

If an enterprise has enough IPv4 addresses for the foreseeable future, it is only guaranteed to be able to talk to itself. An IPv4-only enterprise will not be able to reach services in the cloud that are accessible via IPv6 only (and that will happen).

It's true that most websites have not migrated to IPv6 yet. But this is in part because end users have not migrated -- creating a chicken-and-egg situation. If this cycle is not broken on the content-provider side, then some rather unfortunate long-term consequences for the Internet will arise. Service and content providers will continue to maintain hacked IPv4 implementations involving multiple layers of increasingly dysfunctional NAT, increasing the cost and complexity of Internet connectivity.

Content providers that regard the lack of IPv6-capable clients as a reason to avoid deploying IPv6 for their content are not only jeopardizing their own futures, they're also contributing to the lack of deployed IPv6 capable clients. Until a critical mass of content is available, ISPs must provide IPv4 services to as many of their clients as possible, even if it means degrading those services and increasing costs.

Eventually the costs will become unsustainable and ISPs will pass the costs along to end users, eventually pushing users to IPv6 clients. Unfortunately for content providers, once users leave, they rarely come back. Content providers must be ready for IPv6 before their customers are.

For now, the only non-fear-inducing bet is dual-stack.

— Owen DeLong, IPv6 Evangelist, Hurricane Electric

About the Author(s)

Owen DeLong

Owen DeLong is anIPv6 Evangelist at Hurricane Electric and a member of the ARIN Advisory Council. Owen brings more than 25 years of industry experience. He is an active member of the systems administration, operations, and IP Policy communities. In the past, Owen has worked at Tellme Networks (Senior Network Engineer), Exodus Communications (Senior Backbone Engineer) where he was part of the team that took Exodus from a pre-IPO start up with 2 data centers to a major global provider of hosting services, Netcom Online (Network Engineer) where he worked on a team that moved the internet from an expensive R&E tool to a widely available public access system accessible to anyone with a computer, Sun Microsystems (Senior Systems Administrator), and more.

Subscribe and receive the latest news from the industry.
Join 62,000+ members. Yes it's completely free.

You May Also Like