Momentum for IPv6 services has been steadily gathering. However, whether this is good or bad news depends on your perspective.

And speaking of perspective, there are two things you should know about mine: First, I'm a charter member of the IPv6-haters mailing list. Second, my company builds networks for service providers - and we've been more than happy to work on IPv6 projects.

I'm no IPv6 fan, but I will take money to build an IPv6 network.

IPv6 deployment is more challenging than anyone, including its boosters, ever envisioned, and deployment challenges for service providers will translate to service problems for users.

Most of the service providers I've worked with are doing IPv6 not because they particularly want to, but because they have to. Their networks, or parts of them, are mobile, and handset vendors, such as Ericsson and Nokia, are hard-coding IPv6 support in their devices. Refusing to support IPv6 isn't an option for these service providers.

Yet, service providers are just beginning to discover the difficulty of transitioning a network to IPv6. Check out the Internet Engineering Task Force draft written by engineers at Deutsche Telekom, British Telecom, Telenor and others.

My favorite quote in the draft is: "The current variety of transition mechanisms and [their] potential problems can hinder the deployment of these mechanisms in a large scale." That's engineer-speak for "Dude, this stuff is hard!"

The details of the problems are beyond the scope of this discussion, but they boil down to the multiplicity of options that exist for tunneling and gatewaying between IPv4 and IPv6.

That brings us to the issue of user impact. Anything that causes headaches for service providers potentially causes problems for users. Your local telephone company will swear on a stack of Bibles that nothing will disrupt service to you, the valued customer. . . . Believe that?

The problems are in three categories:

Service disruption -- Making any modification to an operational network is always challenging. Transitioning from IPv4 to IPv6 is like rebuilding a car engine while the car is going 100 mph.

Performance degradation -- Any time you add gateway, translation and brokering services, you potentially increase network latency and slow the performance of customer applications.

Longer provisioning times -- Coming up to speed on IPv6 will reduce the time engineers can spend doing routine operations and architecture. This will elongate service deployment timeframes for IPv6 and IPv4 services.

What's a user to do? Negotiate effective service contracts with well-defined terms, and stringent penalty clauses. That won't fix the underlying complexities, but it will help ensure that your issues get bumped to the top of the carrier's priority list.

» posted by ITworld staff

Network World

• Email
• Print
• Share
  • Slashdot
  • Digg
  • Stumble
  • Reddit
  • Newsvine