The Transition from IPv4 to IPv6

By Jeff Flanagan | June 18, 2013

I just returned from a conference that was recently held in New Orleans, LA. All in all, it was an excellent show featuring information on the latest advanced networking techniques and applications, but one topic that stood out was the looming transition from the IPv4 Data Center 9protocol to the IPv6 protocol.

For those not familiar with this, IPv4 is the primary protocol used on the Internet. IPv4 has only a little more than 32 million of 4 billion addresses (2^32) left in this space. Sure, those are big numbers – but it is predicted that the remaining address space will be allocated (depleted) by 1Q14. IPv6 on the other hand, which was developed about 15 years ago, has 2^128 addresses. To put this in perspective, a typical IPv6 subnet size is larger than the entire IPv4 address space.

Technical Hurdles

The main hurdle here is that IPv6 is not directly compatible with IPv4. Thus, until everyone is using IPv6, systems deployed using IPv6 require backwards compatibility measures – such as IPv6-to-IPv4 gateways – so they can be accessed by IPv4 clients. These gateways are complex and expensive – which has slowed adoption.

Cost is a factor in terms of equipment, services and the labor required to make the transition. There are some technical challenges as well, but in my opinion most companies outside of service providers are really just deferring until they have a need for more IPv4 addresses and find that space is exhausted.

What to Do

There are a few things worth noting about this impending event. First, when all of the IPv4 space is depleted, despite claims to the contrary, the Internet will not stop working. There is no Y2K FUD at play here.

What it does mean is that those organizations with the biggest demand for additional address space, such as ISPs, web hosts, or cloud providers, will not be able to obtain additional addresses when their allotment runs dry, meaning they will need to immediately move to IPv6. Organizations that already have one or more IPv4 blocks assigned to them and that have remaining room in those blocks will be able to continue using them.

It doesn't have to be one or the other or a wait and see approach, though. While there’s no specific deadline or mandate (other than the looming depletion), many organizations that have deployed IPv6 are still running IPv4 in parallel. This is referred to as "dual stack" networking, and allows companies to test IPv6 deployments without having to yet make a full push to IPv6.

Here at Markley, we operate a dual stack network for our Infrastructure-as-a-Service (IaaS) offering, Markley Cloud Services, meaning that customers may choose to deploy their systems using IPv4, IPv6, or both. We’ve found that this has helped our customers address this potential issue without throwing dollars and effort toward something they may not need right now.

Interested in knowing more about this topic? Prefer to make the switch now or to take a dual stack approach? Want to share ideas on how to successfully make the transition – or horror stories of things that have gone wrong? Leave a comment below or tweet to us at our handle at @MarkleyBoston.