APNIC Chief Scientist: IPv6 Is No Longer Important
The network itself is diminishing in the future of Internet architecture
2025-01-17 update
Geoff’s new article offers more expert advice
https://www.potaroo.net/ispcol/2025-01/addr2024.html
Geoff Huston, Chief Scientist of the Asia Pacific Network Information Center (APNIC), pointed out that the root cause of the slow progress of the IPv6 rollout is that the importance of the network itself is diminishing in the future Internet architecture.
IPv6 was once seen as the core technology for the next generation of the Internet but has so far failed to replace IPv4 and has faded from the media and public’s view. Geoff Huston, Chief Scientist at the Asia-Pacific Network Information Center (APNIC), pointed out that the fundamental reason for the slow progress of the IPv6 rollout is that the importance of the “network” itself is diminishing in the future Internet architecture.
From a technical standpoint, Huston argues, IPv6’s stagnation is not a failure of the technology itself, but rather a result of the evolution of the Internet’s architecture and service-delivery model beyond what it was originally designed for.
Why has the rollout of IPv6 fallen short of expectations?
In the early days of IPv6, experts anticipated a natural transition to IPv6 as demand for IPv4 addresses surged. It was even hypothesized that various applications, hosts, and networks would support both IPv6 and IPv4 and then phase out the latter. However, the reality has been different, leading to unexpected challenges in the IPv6 rollout.
Today, while global IPv6 support has reached 40 percent, this is largely driven by countries with shortages of IPv4 addresses, such as China and India. However, this localized success cannot overshadow the general challenges of IPv6 worldwide. The fact that ISPs in many countries and regions have adopted dual-stack (double-stack) technology, a strategy that supports both IPv4 and IPv6, instead of moving entirely to IPv6, is a cause for concern.
In addition, with the rise of the mobile Internet, especially the popularity of smartphones such as the iPhone, network operators’ priorities have shifted to responding to surging subscriber demand. “We either had to focus on expanding network capacity or go ahead and push for IPv6 deployment,” Huston recalled. Operators chose the former, leading to early mobile networks that relied on IPv4 and Network Address Translation (NAT) to connect more devices without the need to assign unique IP addresses to each device.
The advent of NAT and the development of application-layer encryption technologies such as TLS have allowed IPv4 networks to remain viable enough. Since many content providers have not widely adopted IPv6, network operators have no incentive to implement it on a large scale.
IPv6’s biggest adversary: CDNs
Huston points out that IPv6 was originally developed as a response to concerns about IP address exhaustion, but the introduction of IPv6 has not been “revolutionary,” as Huston puts it: “IPv6 doesn’t offer anything new that IPv4 didn’t offer. It doesn’t bring significant changes to the operational aspects of IP, it just extends IP addresses.” In other words, IPv6 is more of a ‘revolution’ than a ‘revolution’. In other words, IPv6 is more of an evolution of IPv4 than a completely redesigned and disruptive technology.
Huston further noted that the rise of content delivery networks (CDNs) may be one of the reasons for the slow rollout of IPv6. CDNs, which are a network of servers distributed geographically, rely on the Domain Name System (DNS) for their operation, rather than IP addresses. This means that the specific version of IP used is less important than the domain name. He emphasized, “From this perspective, IPv4 or IPv6 is not the key resource for services and users; it is the ‘name’ (i.e., domain name) that is the currency of the CDN network.” As a result, IPv6 is not a critical technology in a service delivery environment for large numbers of users.
The network itself is being marginalized?
Huston suggests that the role of the network in the future is also shifting from being a core function of the Internet to being a “simple transport medium,” with applications and services gradually taking over many of the tasks traditionally performed by the network layer. The network layer, in traditional networking, is responsible for routing and forwarding data packets. However, with the evolution of content delivery frameworks and the abundance of computing resources, these tasks are being taken over by applications and services, reducing reliance on specific IP versions. As a result, the role of the network may be further marginalized in the future. “We have progressively stripped core functions from the network, replacing it with a cheap and fast general-purpose data transfer medium, ultimately turning the network into a ‘dumb pipe’.”
He even questioned whether we need to redefine the Internet and whether it should be viewed as a collection of services that share a namespace, and no longer as a network that uses a common protocol for transmission.
