IPV4 to IPV6 Migration, the ‘Real’ Y2K Problem

Failure to migrate to IPV6 will create performance 'speed bumps'

Many believe that warnings about the perils of running out of IPV4 addresses can safely be ignored--that like the Y2K machinations of the last century, they are much ado about nothing. After all, you already have your IPV4 address, so why should you care? Performance, that's why.

Consider this. The coexistence of IPV4 and IPV6 is like having two railroad gauges. Moving cargo across a route with multiple gauges introduces delay to transfer cargo from one train to another before it can continue on its way. In the same way, a network route involving IPV4 and IPV6 "tracks" requires service providers to implement a conversion process that slows performance--and the more conversions along the route, the bigger the performance hit. So, the faster the world transitions from IPV4 to IPV6, the fewer conversion "speed bumps" Internet traffic will ultimately traverse, and the better end user performance will be.

Here's where things stand. John Curran, CEO of the American Registry for Internet Numbers (ARIN), the organization responsible for managing IP addresses for North America, told us over breakfast at Interop that the end is nigh for IPV4 address availability. John is on a mission to spread the word about the exhaustion of the IPV4 address space and the need for a rapid IPV6 adoption. As of this writing ARIN predicts that the world will run out of IPV4 addresses in 13 months. Here is a nifty IPV4 "doomsday" calculator. Depending on when you are reading this posting, the time to doomsday may vary.

At the core of the Internet, tier one ISPs have implemented IPV6. Big swaths of the Internet in developing regions like China adopted IPV6 from the outset. Broadband ISPs like Comcast are preparing to migrate to IPV6, and when they flip the switch some 14 million subscribers will become IPV6 end points. US federal agencies were mandated to upgrade their network backbones to IPv6 by 2008.

Just how big the IPV6/IPV4 conversion performance hit will be remains to be seen, but we can say with confidence that each conversion will add more delay than network address translation (NAT), and that is just an IPV4-to-IPV4 conversion process. Each IPV4-to-IPV6 and IPV6-to-IPV4 conversion adds an extra hop and requires a hefty amount of computing power to process stateful information about who you are trying to talk to and who you are known by on the other end. The more gateway boxes on a route, the more onerous the processing requirements become. So if you have a website, we suggest the sooner you migrate to IPV6 the better because it will minimize the number of conversions your users will have to endure in the long run, and will improve performance.  

Join the Network World communities on Facebook and LinkedIn to comment on topics that are top of mind.
Related:
Now read: Getting grounded in IoT