What Happened to IPv4 Address Space?

What Happened to IPv4 Address Space?

Posted March 30, 2011

1 comment

The original Internet Protocol version 4 (IPv4) protocol was in gradual deployment since the mid-1980s, designed to accommodate 4.3 billion unique addresses (232) that at the time looked like infinity. The real boom in IP allocations started in 1993 as a result of a significant growth in personal computers and continued to escalate due to the explosive growth in mobile devices starting in 2005. This surpassed the wildest growth predictions, leaving the Internet with less than 5% of unallocated address space. Out of the total 256 blocks, as of Aug 2010 only twelve /8 blocks with 16.7m addresses each are remaining. And the rate of address grab is accelerating as we are nearing the depletion of the remaining unallocated IPv4 address space, making IPv4 a victim of its own success.

Is the Internet running out of room? Have we reached the end of the web? Don’t panic, although the Internet Protocol Version 4 (IPv4) is running out of room, back-up plans are in the works.

What’s Next: The Address Solution

Figure 1: Available IPv4 /8 Blocks (as of Aug 2010, in just two months, four more /8 blocks are gone). Graphic source: ARIN.

Over the years, a solution has been devised and widely implemented to extend the capacity of the IPv4 space. Private IP ranges, on internal corporate and residential networks based on the RFC 1918 specification, often now hide behind Network Address and Port Translation (NAPT or NAT44) gateways. These networks share a single or few public IP address for a larger number of users. While this has allowed a significant stretch in the longevity of IPv4, the solution has run its course and is now one of the complications on the road to IPv6.

The ultimate goal is to migrate the internet to IPv6, a protocol with a four times larger address field. For the foreseeable future, this new space is effectively an infinite of space (2128 or IPv4’s 232 x 296). IPv6 has been standardized in 1998 and was in gradually expanding but still very limited deployments since 1999.

Early IPv6 adopters are mostly governments and a few leading-edge multinationals. They are mainly outside of North America, where obtaining IP allocations has always been more of a challenge. While most vendors have offered IPv6 support for nearly a decade, solutions around IPv6 are still not fully mature. What further complicates the migration to a pure IPv6 is that it isn’t backwards compatible with IPv4 and requires significant infrastructure upgrades and support staff training.

Is it all IP gloom and doom?

The official numbers and the associated media frenzy are deceiving. Current IP space allocations are based on a 12-18 months of inventory remaining in each of the 5 Regional Internet Registries (RIRs) around the world—they are responsible for IP assignment to regional ISPs and large enterprises. ISPs are starting to issue private instead of public IP addresses to mobile devices and using LSN (Large Scale Network Address Translation) to handle hordes of mobile users. Small inventories of spare addresses are maintained by the ISPs as well as many governments and enterprises which will allow them to keep expanding even if new allocations are no longer possible.

There is also a black market developing that is turning IPv4 address space ownership into a valuable commodity—Ipv4 space is even considered as an asset during acquisitions. And with an abundance of unused IPv4 address space held by many organizations, the Number Resource Organization (NRO) along with its five member Regional Internet Registries would be smart to turn this into a white market instead and allow trading in unused IPs. This would buy the industry more time on IPv4, but as with any other commodity, will most certainly raise the cost of doing business online.

How do we rid ourselves of IPv4?

IPv6-based Internet is already out there as a fairly isolated island with a limited number of services. Most Internet users and content providers are not connected to it and would depend on their ISPs and changes to their own networks to establish such connectivity.

Figure 2: Multitude of connectivity combinations (Graphic source: Techfuel)

The road to an IPv6-only world starts with tunnelling and translation between IPv4 and IPv6, leading to dual-stack IPv4/IPv6 coexistence, and eventually reaching the ultimate goal of an IPv6-only internet.
During this interim phase, any organization’s content (e.g., websites, applications, email, etc.) must be reachable via both IPv4 and IPv6. Until all potential consumers of company’s content have been fully migrated to IPv6,  interim IPv4-to-IPv6 coexistence solutions are required.

Unfortunately, we can’t reboot the Internet and have it come back up as a native IPv6 network. The migration to IPv6 requires not just significant training, but also upgrades to DNS, operating systems, applications, routers, firewalls, VPNs, load-balancers, broadband CE devices, mobile devices, as well as service provisioning/management applications. It will likely take a decade or two to complete the worldwide migration to a pure IPv6 Internet.

In the interim, the industry developed a number of new technologies to stretch the coexistence of the old and the new while allowing the Internet to continue its growth once IPv4 is depleted. None of these stop gap measures are easy to deploy, but years of address space mismanagement and industry wide failure to implement dual-stack IPv4+IPv6 networks have left the industry with a few imperfect choices. While this transition is happening, there will be two Internets, one based on IPv4 and the other based on IPv6, with a multitude of tunnelling and translation solutions in-between.


These transitional solutions are primarily designed to hide the huge numbers of broadband and mobile users utilizing IPv4 addresses. Solutions like protocol translation and tunnelling, including NAT46, NAT64, NAT464, DS-Lite and LSN are far from perfect, and each introduces its own challenges. This variety of options will also result in market segmentation, with operators choosing what works best for them until the market has moved through the tunnelling/translation and coexistence (dual-stack) phases and reached a uniform IPv6-only world.

Corporations should immediately start planning and preparations through discussions with their network operators, and by initiating internal planning activities, including a review of all systems, applications, networks and security systems for IPv6 transition readiness.


Jakov Zaidman
October 2010

Share this

Recent Comments

Suzette Martin

4/1/2011 2:21:05 PM



Where can I find additional references on this subject?


Post a Comment

Comments on this article are pre-moderated. Before they appear, comments are reviewed by moderators to ensure they meet our guidelines.

Enter security code:
 Security code

About the Author


Jakov Zaidman

Security is one of T4G's main service offerings and, as a Principal Consultant, Jakov is responsible for the networking and security of many of T4G's clients.