The massive, far-reaching, and rapid growth of the internet internationally has undeniably spurred a great deal of innovation and afforded both businesses and individuals a wealth of benefits and convenience. However, as the internet grows and becomes more engrained in people's lives and the day to day operations of organisations from almost every industry, a fundamental and increasingly significant issue has risen to the fore around the sheer number of internet facing devices that are being employed on a wide-scale (particularly those at the network's edge).
This issue, as you may have guessed from the headline, is around the structural limitations of IPv4, the most prevalent generation of the Internet Protocol (of IP address fame) that is essential for routing traffic on the internet. Principally, IPv4 is limited in that there aren't enough IP addresses within the protocol to go around, with the total number of IPv4 addresses now finally reaching a point of actual depletion.
While the logical resolution to this issue - to upgrade from IPv4 to the much more future proof next generation of the protocol IPv6 - has been recognised for quite some time, things aren't as easy as they seem. There are a range of issues standing in the way of IPv6 uptake that continue to hold relevance since it was first introduced in the late 90s. As a result, it's still not exactly clear when IPv6 is actually going to go ‘mainstream', and whether IPv4 will ever really die out.
The problem with IPv4
Despite being first introduced in the early 1980s, Internet Protocol Version 4 is still the most widely used version of IP on the internet today. Using IP addresses (that look something like 172.217.22.14), the protocol is used to identify any particular piece of hardware on the internet and is fundamental for routing traffic and thus the function of the internet itself. As a 32-bit protocol, IPv4 contains 232, or 4.29 billion possible IP addresses. While, at the time it was first conceived, this was considered to be more than enough IP capacity, as billions more devices hit the internet, IPv4 is hitting its absolute limits.
While IPv4 ‘exhaustion' has been talked about for years, the problem has become increasingly concerning as the five international bodies responsible for allocating IPv4 addresses, known as Regional Internet Registries (RIRs), have literally ran out of new IPv4 batches. The latest in this saga came in November last year, when the European RIR, RIPE NCC, announced that it had made its final /22 IPv4 allocation and had thus run out of IPv4 addresses. While the registry will continue assigning IPv4 addresses as businesses close or as networks return addresses they no longer need, RIPE NCC noted that these small allocations, "will not come close to the many millions of addresses that networks in our region need today."
RIPE NCC noted that the development is a further indication that IPv6 should be implemented with a degree of urgency. In a statement announcing the news, the registry said, "Without wide-scale IPv6 deployment, we risk heading into a future where the growth of our Internet is unnecessarily limited - not by a lack of skilled network engineers, technical equipment or investment - but by a shortage of unique network identifiers. There is still a long way to go, and we call on all stakeholders to play their role in supporting the IPv6 roll-out".
Benefits of IPv6
As a 128-bit alternative to IPv4, IPv6 has roughly 340 undecillion (i.e. 340 followed by 36 zeros) possible addresses for internet facing devices. While this number has been contested in terms of the actual, allocable capacity of IPv6, the point remains that there is a lot of IPv6 addresses to the point where the protocol is likely to remain future proof for years to come. An IPv6 address consists of 8 numbered strings separated by colons and looks something like this; 2001:0db8:85a3:0000:0000:8a2e:0370:7334.
However, IPv6 has a string of other benefits including more efficient routing through the size reduction of routing tables, more efficient packet processing, simplified network configuration as well as some security advantages, particularly through the IPSec security protocol. IPv6 network environments can also operate without a traditional DHCP server (the protocol that allows network administrators to assign IP addresses) due to its compatibility with a more lightweight alternative known as SLAAC.
Apple has told its developers to switch over to IPv6 due to speed advantages, citing IPv6 as 1.4 times faster than IPv4 due to improved routing and Network Address Translation (NAT) usage.
Adoption challenges
While this all might paint a picture of IPv6 superiority, unfortunately there are still significant issues that plague the potential of IPv6 to fully take over from IPv4. According to the Internet Society, a major problem that inhibited the transition to the newer protocol was that it didn't offer network operators, enterprises, or vendors any clear advantages in the short term and required a level of expenditure that couldn't be justified as a result. Cost is a real roadblock here, as IPv6 support requires extensive investment in bare-metal and other infrastructure resources.
This issue is made worse by the major compatibility issues that exist between the two versions, as those with ‘full' IPv4 or IPv6 networking stacks cannot access stacks in the opposing version of the protocol. This has long been a thorn in the side of networking administrators looking to implement IPv6, as unless they were willing to cut themselves off from most of the internet, they would be forced to employ both versions in a ‘dual stack' capacity. This further disincentivises moves towards IPv6 as almost everything on the internet is still accessible with IPv4, as the vast majority of networks must offer IPv4 or else risk cutting themselves off.
The Internet Governance project of the Georgia Institute of Technology highlighted this in a recent post, highlighting three possible scenarios for network operators. The first is remain on IPv4 (do nothing), implement dual stack, or run native IPv6 among compatible parts of the network with tunneling or translation tech to make it compatible with IPv4. While claiming that the 3rd option would be the best for the proliferation of IPv6, they note the costs do not justify the required amount of investment.
"There is no difference in the network benefits obtained in all of the three options; all three approaches gain access to essentially the same Internet. Consequently, one network operator's migration to IPv6 places no pressure on the incentives of other network operators to deploy IPv6. There is also no discernible difference in the Internet service offered via IPv6 and IPv4," the post reads.
Keeping IPv4 afloat with NAT
Another major inhibiting factor for the adoption of IPv6 is the efforts that have been made over the years to elongate the lifespan of IPv4. One of the most notable of these is the implementation of Network Address Translation (NAT), which allows an allotment of common privately used IP addresses to be converted or translated to one publicly facing IP address at the network's edge (with translation done on a router for example).
These private addresses can be shared across two local environments, as long as those environments never interact with each other. This essentially means that multiple devices on a network can communicate using only one public IP address, so enterprises don't have to shell out for swathes of IPv4 addresses for each of their devices.
Essentially, up until this point IPv4 has just been easier and cheaper to maintain, its lifecycle has been increased and the difficulties and costs associated with IPv4 haven't been worth it for network operators. However, as IPv4 exhaustion finally takes hold and the cost of maintaining complicated translation driven IPv4 environments increases, IPv6 will ultimately need to go mainstream.
Where are IPv6 adoption levels at now?
According to Google, who run an ongoing measurement of IPv6 availability among its users, the use of the newer protocol has risen steadily since about 2010. However, as of July 2020, only about a third of users (32%) support IPv6. This is quite an extraordinary figure considering the length of time IPv4 exhaustion has been a known issue and paints a bleak picture when it comes those wishing for a time when fully native IPv6 implementation will take over.
The levels of adoption also vary quite considerably country to country, with Germany (47.98%), Greece (46.52%), India (45.65%), Malaysia (44.68%) and the United States (42.47%) leading the charge. Overall, adoption seems to be more prevalent in developed regions, with those in Europe, North America, and Australia all posting relatively high adoption figures (although South America and South East Asia are also posting healthy numbers).
Within the US, a push is also being made on a governmental level to ensure federal organisations start moving to IPv6. The Office of Management and Budget (OMB) published some ‘updated guidance' on the topic in March, providing year-by-year deadlines that ultimately set a goal of 80% IPv6 deployment by 2025. A memo from Russel Vought, who was acting OMB director at the time, cited recent private sector pushes towards IPv6 as the main driver for the federal government to catch up.
However, a large proportion of the world's most visited websites persist without IPv6 compatibility. According to whynoipv6.com (a site that actively monitors IPv6 compatibility), only 382 sites have IPv6 enabled. Such sites neglecting the protocol include Twitter, Reddit, Amazon.com, Baidu, and eBay.
When will IPv4 finally die?
While there is no ‘shut-off' date for IPv4, there have been many predictions as to when IPv6 will fully takeover from IPv4 over the last decade, with some claiming it to be around 2035 while others have said it might take much longer. The truth is, though, it's still quite difficult to determine both when IPv6 will actually be considered the main protocol and when (or even if) IPv4 will completely die out.
According to Vincentas Grinius, CEO at IP leasing company Heficed, the lack of obvious upgrade incentives is an issue that continues to plague the growth of IPv6, with relatively few volunteers willing to act on the frontlines of the transition.
"Without a doubt, IPv6's maturity is increasing, but there's still a long road ahead before it is as fully accessible as it needs to be, in order to facilitate the Network's further expansion," Grinius explains. "It could be years till the majority of tech runs only with version 6, and even then, most of the hardware will need to have backlink compatibility, as there will be older phones, laptops, or other devices that need IPv4 to connect."
However, the good news for IPv6 is that companies and network operators who are reliant upon fast customer and network growth will inevitably have to support IPv6, as the IPv4 exhaustion really starts to take hold. There are also a couple of technological trends that will surely keep IPv6 growth in an upwards trajectory, not the least of which is large-scale IoT deployments.
As IPv6 assigns specific IP addresses to every internet facing device (instead of relying on private IP translation with NAT), businesses who deploy IoT will be able to effectively do away with NAT and thus simplify their networking device stacks. This would also lead to some cost savings, as maintaining NAT devices can be costly and complicated with the more devices on the network. Additionally, as IoT devices hit the tens of billions, even NAT won't be able to save these ecosystems from exhausting their IPv4 options. IPv4 also has inefficiencies when it comes to preserving battery life due to a decrease in broadcasting necessities.
Ultimately, it's a bit of a mixed bag for IPv6. While its adoption will continue to rise, it may continue to do so at a snail's pace due to the historical issues that are still relevant to this day. While it's equal parts crazy and fascinating that a protocol developed over 20 years ago, in the world of technology, would still be considered as the rising "next generation", it seems that, at least for the moment, that theme is set to continue.