OTTAWA — Mark Blanchet, president of Viagénie Inc., says there are two types of companies when it comes to IPv6 in Canada: the turtles and the rabbits.
Perhaps not surprisingly, the turtles, who prodded along taking small, deliberate steps to deal with the Internet addressing problem, are much closer to the finish line. The rabbits, who dillydallied believing they had plenty of time, are now sprinting, trying to make up ground.
“The turtles are the ones that have been moving and are roughly ready,” Blanchet told a crowd at the University of Ottawa last week who attended the IPv6 Summit.
The rabbits, on the other hand are “the ones that have been waiting, not doing anything and now are trying to catch up,” he added during his presentation entitled, The Turtle and the Rabbit: Perspectives on IPv6 deployment in Canada.
The status of IPv6 deployment in Canada is a mixed bag with many large ISPs having laid some of the ground work. Telco-based providers, who have multi-protocol label switching (MPLS) networks, shouldn’t have too much problem, according to Blanchet. The cable operators are also fairly well prepared, at least those who are rolling out DOCSIS 3.0-based services, he said. In the mobile arena, an initiative by T-Mobile in the U.S. could be a sign of what’s to come to the Canadian mobile wireless market. The U.S. wireless operator is considering an IPv6-only network because 80% of the data traffic goes to a few content providers (Google, Facebook and Yahoo) which are already IPv6 ready.
Other organizations such as the Communications Research Centre, the National Research Council of Canada, CANARIE, Tata Communications, the Canadian Internet Registration Authority and others can be considered turtles. Interestingly, the Quebec Government is ahead of the game as well.
In 2008 when it issued a request for proposal for government network services, it included a provision for IPv6, according to Blanchet. The winning bidder had to be able to implement IPv6 when the Quebec Government was ready. Telus Corp. won that contract.
Getting to IPv6 isn’t a simple process. One solution is to adopt an IP address sharing strategy, but that has negative impacts on everything from spam to tracking for law enforcement and online marketing to applications that require multiple simultaneous connections.
Because there could be 5,000 users behind a single IP address, whitelisting and blacklisting for spam won’t work because all users behind that single IP address will be affected, said Blanchet. Equally, the use of geo-location in online marketing activities will be less useable and “logging and tracking of end users for copyright [and] law enforcement doesn’t work because there’s no way to easily identify the users,” he added.
Owen DeLong, an IPv6 evangelist at Hurricane Electric, highlighted many of the same issues in his presentation, IPv6 — Light at the End of the Tunnel. He noted the light could either be all the hard work coming to fruition or an oncoming train that’s going to require a complete reboot of the corporate communications network.
“Overnight you need to deploy an entirely new protocol in your network and you have no idea how to get there from where you are,” he said.
DeLong offered plenty of advice regarding the benefits of IPv6. He added though that this shouldn’t be seen as a major undertaking and a complete swap out of old IPv4 infrastructure for newer, more advanced IPv6-enabled gear. “We’re simply talking about leaving your existing IPv4 network in place working as it is, and adding IPv6 capabilities on top of that,” he said during his keynote. “That’s not such a major change. It’s not completely minor, but it’s not as big as a lot of people seem to think.”
Another problem that has led to a delay in IPv6 deployments is the lack of a business case. DeLong said there is no “killer app” and no real return on investment, adding the rationale for doing IPv6 should be the same as a disaster recovery strategy and Y2K planning.
“If you don’t have IPv6,” he said, “you will be increasingly disadvantaged compared to your competitors that do. You will become increasingly disconnected from the majority of the internet.”
As with most new technology deployments, start small and target the low-hanging fruit. The test lab could include a deployment of IPv6 at the peering edges and then move it across the corporate system in a contiguous manner. Add public content such as e-mail and the Web, and to enterprise systems where it makes sense.
For a test lab, simulating the entre internet isn’t required, DeLong said. A few routers and test systems will suffice because it’s about testing network configuration.
“Beginning your implementation now allows you to do a somewhat slower, safe progression towards full integration in a controlled manner. You can plan your spending, you can research which products meet your needs best, you can negotiate with your vendors on pricing,” he explained.
Otherwise, he added, you may find yourself racing to the finish line, doing shoddy work, paying more than you need to and losing customers in the process.
