We live in times when technology is changing human activities at an unprecedented speed. Along this journey, the greatest opportunity —and at the same time the greatest challenge— has been the transfer of many of these activities to the Internet. This is why any change in its structure and operation has significant impacts.
Why act at the corporate strategy level?
Companies need to undertake digital transformation to capture opportunities and survive. However, we rarely reflect on the fact that the greatest impact has not actually been produced by the use of digital devices (personal computers, tablets, mobile phones, TV, software applications…) in isolation (20th century), but by the interconnection of all of these devices to create a global village based on a network of networks known as the Internet.
The Internet’s success was the result of its open, cooperative model and royalty-free standards, which continue to be defined by the Internet Engineering Task Force (IETF), an organization that operates under that same open model.
The Internet was designed as an interconnection network for the defense sector and for universities and research centers to share knowledge, not for the crucial role of hosting the transformation of every activity and sector.
Proof of this is the limitation on IP addressing (the identifiers we assign to each element that connects to the network), yet this is not the only limitation that must be solved. Defining an extensible standard was imperative. Because of this, the IETF defined the extensible and scalable IPv6 standard as early as 1998 to replace the limited IPv4 standard, which dates back to 1983.
(Free access, no subscription required)
This change has taken almost 20 years. Now, the global IPv6 deployment level has already reached 40%, and in the large digital economies this level exceeds or is about to exceed 50%.
Following the adoption of IPv6 by tech giants (Google, Facebook, Apple, Amazon, Netflix, Uber, Spotify, Microsoft, etc.) and Internet providers serving end users, the next point in the value chain are companies that are already conducting their commercial activities on the Internet.
Why act now, in 2023?
This change has taken almost 20 years. Now, the global IPv6 deployment level has already reached 40%, and in the large digital economies this level exceeds or is about to exceed 50%.
Following the adoption of IPv6 by tech giants (Google, Facebook, Apple, Amazon, Netflix, Uber, Spotify, Microsoft, etc.) and Internet providers serving end users, the next point in the value chain are companies that are already conducting their commercial activities on the Internet.
Why act now, in 2023?
In its prestigious 2021 analysis of network technologies in corporate environments, Gartner estimated that IPv6 is a mature technology in the “Slope of Enlightenment” and that it will be adopted by the majority between 2026-2031 (5 to 10 years after the study).
Not adopting IPv6 means not understanding and not putting to test the challenges and opportunities of the emerging Internet. As the examples below will show, the tech giants have already been down part of this road.
A very recent example involves Apple, Google, Amazon, and Samsung, who have based their Matter/Thread alliance for IoT for the home on the IETF 6LowPAN standard for low power networks (LowPANs), which has no equivalent in IPv4. This standard is being adopted by hundreds of companies in this sector. Thus, mesh networks of home devices use IPv6 only among themselves, even if they have to communicate with the Internet or company platforms using IPv4.
Sources: Matter software development kit (SDK) from GitHub. / CNET
Another specific example is the SRv6(Segment Routing IPv6)traffic engineering model for Internet WAN networks which has been standardized by the IETF but has no equivalent for IPv4 networks, although it does for specific networks below the IP level (SR-MPLS). Multinational companies that are migrating from WAN models based on MPLS networks to models with a greater reliance on Internet links or which want to take segmentation to an end-to-end level (up to the Datacenter host), have a good opportunity with SRv6.
The last example of opportunities we will share today is that of a part of the Web3.0 community that supports a much more distributed Blockchain model for transactions which is deeply rooted in the design and use of the IPv6 standard. We do not know whether this option will crystalize in the Web3.0 or metaverse ecosystems, but it would be risky not to understand it and not be ready to get involved in it if it does go forward.
In addition, not adopting IPv6 means ignoring that there are software applications, systems, licenses, firmware, and devices that may not be supported, may not operate optimally, or may generate extra costs in IPv6-only environments that the Chinese government and American agencies have identified for 2030.
How to be efficient in the corporate adoption of IPv6
The most important thing is to understand the issues that the non-adoption of IPv6 is causing in terms of efficiency. This is connected to the issues that companies are seeing emerge in various areas:
In the Cloud, AWS has identified that the backends of containerized services (for example, services with Kubernetes) are difficult to scale due to the scarcity and complexity of private IPv4 domain interconnection.
In the field of mass-market applications and services, tech giants such as Apple or Netflix have noted that the deployment of IPv6-only and IPv4-as-a-service in mobile networks (464xLAT technology) in the United States, India, and China is increasing perceived quality or end-to-end user experience (UX) for their services.
In IoT for utilities, many public tenders for water, electricity or gas meters already require IPv6-only cellular connectivity to avoid deploying thousands of private-to-public address translators (NAT44).
In IoT for the home, the Matter/Thread standard assumes that millions of devices capable of bidirectional end-to-end connection do so natively over IPv6, which means that IPv4 platforms have a harder time leveraging this model (for example, with protocols such as CoAP over UDP, which is more efficient and interactive than HTTP over TCP).
In industrial IoT, the mass connectivity of private 5G or Wi-Fi6 networks and complex solutions such as digital twins or immersive experiences will require simplified network architectures without points of failure or bottlenecks such as network address translators (NAT).
In network engineering, large companies must interconnect private IPv4 address domains avoiding collisions, especially in the case of internal reorganizations or mergers and acquisitions. The internal use of unauthorized public addressing is often resorted to, which means that resources are needed to isolate the devices that utilize those public addresses and prevent external leaks.
As for regulation, government agencies and authorities in the United States, China, Hispanic America, India or the European Union are aware of the change in model and have been actively promoting the adoption of IPv6, but they are already beginning to consider deadlines and strategies to disable IPv4 (IPv6-only and IPv4 sunsetting models). This is especially important for companies that operate in critical infrastructure sectors or regulated businesses.
Once a company is aware that efficiency involves the adoption of IPv6, it is very important that they learn from the experience of ISPs and Internet giants so that this evolution will also be an economically efficient process. For this reason, it is important to consider the following recommendations:
IPv6 adoption is a strategic and crosscutting process, in which a small internal group must initially be involved, but where areas such as networks, security, platforms, systems, backend/frontend, datacenter/cloud, marketing (user experience and applications/services), finances, and purchases must be represented. This will allow planning and defining priorities, but above all, coordinating with suppliers and avoiding missteps.
Training is essential and will keep us from having to invest in external help when time is of the essence or when problems are encountered.
Technical departments will feel greater security adopting IPv6 in parallel, in other words, using the dual-stack strategy (IPv4+IPv6), but it has been shown that this can be quite expensive for ISPs. It seems that the most reasonable strategy is to use IPv6-only wherever possible (greenfield and non-critical brownfield deployments) and dual-stack where there is no other choice (critical brownfield deployments), but always with an IPv4-as-a-service plan or even IPv4 sunsetting to save/optimize costs.
Although the service will initially be dual-stack, setting up IPv6-only environments (IPv6-only Sandbox) is often the only strategy that allows identifying which external and internal suppliers are actually making progress and to analyze failures or vulnerabilities. This is important to avoid problems and additional costs due to breaches and unforeseen breakdowns in the company’s value chain.
IPv6 Council Spain looks forward to your suggestions or comments, which you can send to: ipv6forum.spain@gmail.com
PS: In a future article we will address IPv6 in the Enterprise: How to Explore Business Opportunities
