We wrote last month of the impending World IPv6 Day on June 8th, which came and went without much fanfare. More than 400 websites and Internet service providers (ISPs) around the world, including Facebook, Google, Yahoo, Akamai, and Limelight Networks, participated in the 24-hour event that provided a coordinated “test flight” of their IPv6-enabled services. Some reports of the event state that there were a few glitches, but no major catastrophes. So can we take the relative silence that ensued after the test, as a town crier’s “all is well” call?

Almost 1 million smartphones shipped in Q1 of 2011… each of these devices require a unique address. The IANA ran out of allocatable IPv4 address blocks on Feb 1. The regional registries are projected to run out of IPv4 by the end of year… Only 50 percent of service providers have enabled IPv6 right now. If all is well, why did Microsoft recently purchase 666,624 IPv4 addresses from Nortel’s liquidation sale for $7.5 million? That’s $11.25USD for each IPv4 address that the IANA was giving away for free until they ran out!

The stakes are high for getting the transition to IPv6 right, and the necessity is fast-approaching. Despite headlines such as “Successful World IPv6 Day Demonstrates Global Readiness for IPv6”, we need to be cautious to prevent misinterpreting the results of World IPv6 day, as the data was not complete.

1. Ill-prepared providers not represented. There are a large number of access providers, websites, and ISPs who did not participate in the event because they are not yet equipped to support IPv6. Some backbone providers still need to establish IPv6 peering with each other. We cannot draw any conclusion that ‘all is well’ for those that did not participate.
2. Scale of use was limited. Because the number of users accessing IPv6 content is currently limited, the test didn’t push any scalability or processing demands. Data from a recent Arbor Networks study suggests that IPv6 currently makes up only between 0.1 and 0.2 percent of Internet traffic. In addition, many access networks have not prepared or tested at scale. This is still a big concern as more and more IPv6 endpoints connect to networks and as more users try to access IPv6 content and applications.
3. Testing time too narrow. World IPv6 day was just a 24-hour test, when a small fraction of websites, most using the dual-stack protocol to enable IPv4/IPv6 coexistence, discovered their readiness for IPv6 connectivity and content. A 24-hour test period is not sufficient to flush-out problems that will occur with a long-term turn-up.
4. Lack of detailed data. Information on the results of World IPv6 day are lacking regarding the collection of results and metrics. The only statement on the website of the promoters of the event, the Internet Society (ISOC), is: “The vast majority of users were able to access services as usual, but in rare cases, users experienced impaired access to participating websites during the trial.” What were the problems users experienced? Was there any way to identify the cause of the problems, such as a point of failure or responsible party? How were problems resolved? We were unable to locate a central repository of findings. A few participants wrote briefs about their World IPv6 day experience, but most details are lacking.

Although we really cannot draw any general conclusions as to ‘world readiness’ for IPv6 based on the event, there is value in that the World IPv6 day drew attention to the need for IPv6 readiness. Web and service providers were given a reality check as to their readiness or lack of readiness to even participate in the event. The next steps are critical for both network equipment manufacturers (NEMs) and service providers.

Test equipment is playing a critical role in accelerating the deployment of IPv6 on broadband access equipment and networks. Ixia is working with NEMs worldwide to ensure their broadband access devices can satisfy IPv6 conformance, interoperability and performance requirements, along with IPv4 co-existence.

For service providers, Ixia models clients, services, and other network equipment in a controlled lab environment so they can emulate different IPv6 deployment scenarios and test them under both realistic and ‘extreme scale’ conditions. They can easily manipulate/modify the test configuration to model different network conditions and future expectations – such as the ratio of IPv4 to IPv6 clients and traffic, and understand how this will impact network performance and end user experience.

With Ixia’s test platform, service providers can achieve repeatable, reliable measurements that enable them to understand scalability and performance thresholds for network configuration purposes, and to ensure they can meet SLAs. By replacing real devices with a few Ixia test ports that effectively emulate an entire network, service providers gain confidence in their IPv6 implementations and can deploy them faster.

Please visit Ixia’s Black Book index to get an IPv6 Transition Technologies Black Book. It includes IPv6 test methodologies (for DSLite, 6rd, Dual Stack PPP, and DHCPv6/PD) that will help Ixia users become familiar with these technologies and understand the key testing challenges associated with them.