IPv6 Update

Companies like Yamaha have had IPv6 capable home routers for over 5 years (in Japan). Linksys WRVS4400N wireless routers have IPv6 support. They work fine, but Linksys business series gear and much more expensive than the normal consumer line.

I see this as an economic decision on the part of the manufacturers like Linksys. There are three primary aspects to their feature prioritization in product development.

Customer Demand – The bulk of North American consumers are not asking for IPv6 capabilities. They do not understand what is will do for them. Part of this lack of awareness is also tied to the lack of IPv6-enabled services. IPv6-based peer-to-peer gaming is possible with Sony PlayStations and the Nintendo WII that have IPv6 on by default. Vista and MAC machines also have IPv6 enabled out of the box.

Carrier Support – IPv6 between home computers on the same layer-2 LAN segment will work using link local addresses. However, carriers delivering broadband to the home in North America are generally NOT providing IPv6 support to support P2P services and interaction between homes. This fact is constraining the demand for IPv6 services in the home. Comcast has started to IPv6-enable their network, starting from the core. It will take a while before they get to the edge. Data Over Cable Service Interface Specification (DOCSIS) 3.0 requires IPv6 support and is being adopted by the industry. The good news is that Ambit, Arris, Cisco, Motorola (2 modems), and SMC have DOCSIS 3.0-certified cable modems. This link may interest you: http://www.cablelabs.com/news/pr/2006/06_pr_docsis30_080706.html

Lack of Global Strategic Vision – Standards required to implement IPv6 in home network equipment have been stable for years. Many companies have a razor-sharp focus on the impact of their decisions on short-term financial results, often ignoring the value of strategic product investments. Having an “IPv6 Ready” logo on the retail package would attract some early adopters, be a potential product differentiator, and also raise the image of the manufacturer as an industry leader. Link: http://www.ipv6ready.org/

Very perceptive, and progress is already underway. The 3rd Generation Partnership Project (3GPP) started almost 10 years ago to address emerging 3G technology standards. The common interests of the members of this industry organization include the development of standards that will support new interoperable services for 3G and beyond. A cornerstone of their services model is the IP Multimedia Subsystem (IMS) specification for converged voice data, and video for mobile devices over IP. IPv6 support is a requirement of this important standard. The carriers are looking at the revenue angle “what new services can we sell?” The IMS specification provides a clean solution foundation through the use of IPv6. Link: http://www.3gpp.org/

Another area of IPv6 for mobile carriers is easier of address management move mobile devices. Many carriers have depended on IPv4 private addressing as defined in RFC 1918. Cell towers constantly shuffle limited blocks of private addresses based on end node device activity. The use of IPv6 instead of IPv4 reduces address management burden and also enables new mobility features. Nokia has been a leader in IPv6 for the mobile environment. Microsoft has supported IPv6 in Windows Mobile for the past 5 years, with each version becoming more feature-rich.

The important thing to note is that IPv6 is not the end game, but a required ingredient to enable new services that have commercial value.

As Fred pointed out earlier, home gateway products supporting IPv6 have been available for some time. It is unclear if the support in the home gateways was the gating factor for service providers to offer IPv6 services. It is the eternal IPv6 story, of some chicken and some egg. While taking the strategic perspective on IPv6, manufacturers prioritize based on market demand as well. In the case of Linksys, official IPv6 support will be available shortly.

Good (not stupid) question. The level of effort [pain] is a question that we all ask when facing the unknown. From the perspective of someone who has led the successful IPv6 implementation in a major global enterprise, IPv6 is a lot easier that past network protocol changes we endured. Any yes, I have gone through SNA/SDLC, HDLC, DECnet, IPX, and Vines transitions. Here are some observations

IPv6 built into the HOST OS
Bechtel’s implementation used Windows XP and Windows Server 2003 as primary hosts for enabling IPv6. The process is straight forward enable IPv6, turn off unneeded transition tunnels. With previous technologies, the network protocol was an add-on. IPv6 is on by default in Vista and Windows Server 2008.

Network OS Support
Cisco IOS, JUNOS, and other popular network operating systems have supported IPv6 for a while. Juniper has IPv6 enabled by default on a majority of their products. Cisco does not. In both cases the capabilities are there. Since IPv6 and IPv6 share common ports, port-based rules can apply to both protocols; i.e. block port 80, allow port 443. Routing paradigms are also comfortably parallel and/or converged. OSPFv3 (IPv6) and OSPFv2 (IPv4)

Common Naming Infrastructure
Both IPv4 and IPv6 use DNS as a common naming / locating infrastructure. AAAA resource records for IPv6 complement the “A” resource records for IPv4 allowing a host to be easily located using wither protocol.

Protocol Translation Avoidance
Bechtel’s approach is ubiquitous dual-stack for the next several years. While there may be a slight administrative overhead with this approach, it has removed the need to implement host-based or network-based protocol translators. If you have had networks running DECnet, IPX, SNA, and IPv4 concurrently, you’ll REALLY appreciate the significance of this.

Addressing
IPv6 addresses provide a lot more flexibility, but also require some new thought processes. In the transition from DECnet to TCI/IP, some companies applied the DECnet area paradigm to their new IPv4 network topology. While this may have made their mental transition a bit easier, this was not necessarily a smart strategic move. Implementation of IPv6 in an IPv4 environment will be easier, but there are opportunities with IPv6 addressing that must be understood and implemented early, if possible. IPv6 multicast and Mobile IPv6 are a couple of examples. In 2005 we learned, discussed, and debated Bechtel’s IPv6 numbering plan for months before settling on a final approach.

Stop the Bleeding
One of the many suggestions in our book, Global IPv6 Strategies, is to stop purchasing / developing products that are not IPv6 aware. Perpetuating a dependence on IPv4 in now a wise thing with broader global deployment of IPv6 already starting.