Experts make a solid business case for IPv6

The co-authors of Global IPv6 Strategies were the guests for a live Network World chat where they discussed all things IPv6. The group consisted of Fred Wettling (pictured, top), who manages architecture and strategic planning for Bechtel and is a member of the IEEE North American IPv6 Task Force, the IPv6 Forum, and is executive director of the IPv6 Business Council; Patrick Grossetete (pictured, middle), who is technical director of product management at ArchRock, makers of IP-based wireless sensor network technology; and Ciprian Popoviciu (pictured,bottom) PhD, CCIE No. 4499, who is a technical leader at Cisco Systems. They discussed the business case for IPv6, killer apps, security tools and the role of vendors like Microsoft and Cisco.

Chip_Popoviciu: Hello, happy to be here!

Fred_Wettling: Hi all - glad you joined us today.

Patrick_Grossetete: Hello everybody, glad to be here.

Moderator-Julie: While Fred, Patrick and Chip are typing their answers to your first questions, here's one that was submitted earlier. What is being done to reconcile the profile differences between DoD and NIST? Who is running the store?

Fred_Wettling: The basic issue you raise is that of unambiguous technical specifications that can be used to qualify U.S. government purchases. The process involves mapping use case (i.e. network-enabled war-fighter) to network function (routing) to the relevant IPv6 specification(s). DOD work on developing IPv6 profiles for qualifying IPv6 products was started several years ago by Defense information Systems Agency (DISA). Their use cases were specific to their missions. NIST profiles are targeted at most of the rest of the U.S. government and may also serve as reference models for other industries. DISA and NIST have been collaborating where possible. There is a lot of common ground. However, there are differences that will exist going forward that are being addressed. I understand dialogs are still going on. There seems to be some convergence toward the NIST profile, including a partnership with the IPv6 Read Logo.

Fred_Wettling: (continued) Who’s running the store? It’s DISA for DOD, NIST for most other agencies. Outside of the U.S. federal space, the industry is deciding the standards applicable to IPv6 core functions. Many vendors are clearly mapping IPv6 capabilities to versions of their products. My recommendation is use the method of qualifying products that are most applicable to your environment.

Layer8: RFC1918 provides more than ample IP space for the largest of enterprises, outside of ISPs and service providers. What would be some compelling reasons to adopt IPv6 as an internal IP architecture?

Fred_Wettling: RFC 1918 does not address communications between organizations. The world is moving to end-to-end communications between people, organizations and objects. The number of people and things that are being connected exceeds the number of possible IPv4 addresses (about 4 billion).

Chip_Popoviciu: Yes, RFC 1918 address space was sufficiently large for most of the needs of today's enterprises but that might not be the case going forward. Enterprises have layered services (data, voice, wireless, etc) each with an addressing scheme. These layers will take significant resources. Then we have the new devices coming into the environment -- sensors, readers, security devices. All these laments can make RFC 1918 insufficient for future needs.

Patrick_Grossetete: First, I would say that not everybody wants to share the same address space. Then, even net 10 is not large enough for large organizations if you consider the number of subnets created today for Wi-Fi or IP telephony deployments. If a large organization experiences a merger (very common) it will require re-numbering, which is always painful.

mikes: The government was originally committed to 2008 for IPv6 and now has a date of 2010. How real is this considering many government organizations have not even started? (And IPv6 has 3.4+38 verses IPv4's 4 billion.)

Patrick_Grossetete: I don't think the date by itself is important. It is more about acknowledging there is a lot to be done. Most world regions have IPv6 on their road maps. Japan and Korea having been the first to start.

Fred_Wettling: U.S. agencies have the challenge of quantifying technical specifications as part of their procurement process. The definition of IPv6 profiles has taken longer than anticipated.

Patrick_Grossetete: What's really important is the planning and education phases, which as we know require times. When done, the deployment should not be different from other networking projects. By the way, that means we look at integration and co-existence, not a full transition to start with

Chip_Popoviciu: Setting target dates is very important in getting things going. The 2008 mandate stimulated various work (such as the profiles) and this will continue. Also, note that several governments around the world have aggressive plans in place for integration and migration.