For more than a decade, the internet has been operating under a quiet constraint: the exhaustion of IPv4 addresses.
Officially depleted in most regions since the early 2010s, IPv4 was expected to give way to IPv6 as the next standard for internet addressing. Yet, despite the urgency and the clear technical advantages of IPv6, adoption has been far slower and more uneven than many predicted.
The reality is that IPv4 is not gone. It is still deeply embedded in the infrastructure of the internet, and the transition to IPv6 is proving to be less of a clean switch and more of a prolonged, complex evolution.
IPv4 Is Exhausted, But Not Obsolete
IPv4 uses a 32-bit address space, allowing for approximately 4.3 billion unique addresses. With the rapid growth of internet-connected devices, cloud services, and global connectivity, this pool has effectively been exhausted. Regional Internet Registries (RIRs) have long since allocated their final blocks.
However, exhaustion does not mean disappearance. IPv4 addresses are still actively used, traded, and repurposed. A secondary market has emerged where organizations buy and sell IPv4 address blocks, sometimes at significant cost. This alone highlights how dependent the modern internet still is on a protocol designed in a very different era.
Instead of replacing IPv4 outright, the industry has found ways to extend its lifespan.
The Rise of Workarounds: CGNAT and Address Sharing
One of the most common strategies used by internet service providers (ISPs) is Carrier-Grade NAT (CGNAT). This approach allows multiple users to share a single public IPv4 address, effectively multiplying the usability of the remaining address space.
While CGNAT solves the immediate problem of address scarcity, it introduces trade-offs. It can complicate peer-to-peer connections, impact latency-sensitive applications, and limit the ability for users to host services or access certain ports. From a data and infrastructure perspective, it adds another layer of complexity to network management and troubleshooting.
More importantly, CGNAT is a workaround, not a long-term solution.
IPv6: The Technically Superior Successor
IPv6 was designed to solve the limitations of IPv4 by introducing a 128-bit address space. This allows for an almost unimaginably large number of unique addresses, enough to support the continued expansion of the internet for decades to come.
Beyond address capacity, IPv6 offers several technical improvements, including more efficient routing, simplified packet headers, and built-in support for modern networking requirements. In theory, the transition should be an obvious choice.
In practice, it is not that simple.
Adoption Is Uneven and Fragmented
Global IPv6 adoption has been steadily increasing, but the distribution is highly uneven. Some countries and networks have embraced IPv6 at scale, while others remain heavily reliant on IPv4.
Large technology companies and content providers have largely enabled IPv6 across their platforms. Major operating systems and devices support it by default. Yet the bottleneck often lies elsewhere in the ecosystem.
Tomas Novosad, founder of Fiber At My Address, notes that while IPv6 support is increasing, many ISPs still operate dual-stack or even IPv4-heavy networks due to legacy infrastructure and customer equipment limitations.
This dual-stack approach, where networks run both IPv4 and IPv6 simultaneously, has become the dominant transition strategy. It allows compatibility with existing systems while gradually introducing IPv6. However, it also slows down full migration, as IPv4 remains necessary for a significant portion of traffic.
Why ISPs Are the Bottleneck
The role of ISPs in the transition to IPv6 cannot be overstated. While enabling IPv6 at the network level is technically feasible, deploying it at scale involves significant operational and financial challenges.
First, there is the issue of infrastructure. Many ISP networks were built and optimized around IPv4. Upgrading core systems, provisioning tools, and monitoring platforms to fully support IPv6 requires time and investment.
Second, there is the customer equipment layer. Millions of routers, modems, and embedded devices in homes and businesses may not fully support IPv6 or may require firmware updates. Coordinating upgrades across such a distributed environment is a non-trivial task.
Third, there is a lack of immediate economic incentive. IPv4, despite its limitations, still works. Through techniques like CGNAT and address trading, ISPs can continue operating without urgent pressure to migrate. As a result, IPv6 adoption often becomes a gradual, low-priority initiative rather than a strategic overhaul.
The Data Perspective: A Long Transition Ahead
From a data standpoint, the transition to IPv6 is not a binary event but a long-term trend. Metrics from various network measurement platforms show steady growth in IPv6 traffic, but also confirm the persistence of IPv4.
This coexistence introduces complexity for data analysis, network optimization, and performance measurement. Systems must account for differences in routing, latency, and connectivity between the two protocols. For businesses relying on accurate network data, understanding this dual environment is increasingly important.
Conclusion
The story of IPv6 adoption is not one of failure, but of gradual evolution constrained by real-world infrastructure. IPv4 may be exhausted, but it remains a foundational part of the internet. IPv6 is the future, but that future is arriving slowly and unevenly.
For now, the internet runs on both. And until the economic, technical, and operational barriers are fully addressed, that dual reality is likely to persist far longer than originally expected.