It’s a bit like coastal erosion – the owners of houses on a cliff know that one day their homes will fall into the sea, but because it isn’t imminent, and no-one can tell them exactly when it will happen, they can continue as if nothing is wrong, living in the ‘now’; or they can make detailed financial and personal preparations to be ready for the day.
The implications of transition from IPv4 to IPv6 are, of course, a lot more complex, but companies who don’t take account of them will one day ‘fall off the cliff’.
Why should we care about transition?
Undoubtedly the UK has long been at the forefront of technological innovation and we’ve managed to stay ahead of most of our competitors simply by moving with the times, setting trends and being quick to react to others’ new ideas. This ability to innovate and retain our global positioning has helped to underpin the UK economy over many years, but the need to transition to IPv6 threatens to de-stabilize this position of influence.
There’s no doubt that the UK has been dragging its heels in drawing up a schedule for transition, for the most part because of the perceived costs involved in running parallel networks, because IPv4 and IPv6 aren’t compatible, but also because of the potential and unresolved security issues that the IPv6 infrastructure will bring. Quantifying the problems in the UK is difficult and attempts to do so have reduced some companies to a state of inaction when faced with the scale of the problem.
But continued inaction is not an option, as the ongoing availability of IP addresses is critical to our ability to connect innovative applications which would otherwise be severely restricted. And, with access to IPv4 addresses rapidly running out, and no prospect of further ones being made available, IPv6 is the only way forward.
IPv6 transition issues
The allocation of IPv4 addresses is based on the number of variations available using a 32-bit address space. When it was conceived, back in the 1970s, the idea that we would ever need more than 4 Billion addresses was unthinkable, but by the 1990s, with an increasing number of applications taking up more than one internet address, it was clear that IPv4 capability wouldn’t be enough.
Many computers already share addresses - giving more capacity on one level - but this merely creates problems by inhibiting the deployment of new applications and services. In any case, with the introduction of intelligent infrastructure for some utilities, healthcare networks, disaster management systems and industrial automation, amongst other examples, all of which require an IP address, the demand is voracious and virtually infinite.
Which is where IPv6 comes in. Rather than using a 32-bit string, IPv6 uses a 128-bit string, giving a much larger number of address variations (think 340 followed by 36 zeros and you’d almost be there!). IPv6 also provides for other improvements in processing and efficiency, although the key interest will be in the larger address space.
So, it’s a ‘no-brainer’ isn’t it?
Not really. Whilst the vastly increased number of addresses, plus other processing and efficiency improvements is the good news, there are some serious downsides – foremost of which is that the two systems are non-compatible, which means that anyone who has switched to IPv6 can’t talk to anyone who is still on IPv4.
There have been a number of ‘work-arounds’ tabled, but the only one that makes sense is to run parallel systems. Although not itself ideal, this option does give users the chance to build an IPv6 infrastructure, test it and run it in parallel, whilst still having the security of IPv4 still running.
How soon will it all happen?
The truth about timescale questions is that no-one really knows. All anyone knows is that at some point it will have to happen and those who make the transition quickest will get a development jump-start on everyone else – how much of a jump-start remains to be seen.
One of the reasons for this lack of clarity is that it is known that there are a lot of IPv4 addresses that have not yet been used, having been stockpiled by larger ISPs and other corporations. How many are still available isn’t known, but estimates of how quickly they will run out vary from between 18 months and 5-10 years.
Another is the use, by some providers, of CGN technology, to prolong the life of the stocks of IPv4 they have. This will certainly help, but doesn’t offer an alternative long-term strategy.
So, rather than be like the cliff-house-dweller who sips his tea in blissful ignorance, the smart thing to do would be to look into parallel-running and at least get a team to do some investigative work on your IPv6 implementation options. That way you can plan for ultimate transition from a position of knowledge, having a better understanding of timescales, costs and any potential disruption to your business.