The IPv6 Conundrum

Written By:
Published:
Content Copyright © 2015 Bloor. All Rights Reserved.
Also posted on: The Norfolk Punt

As is well known, the Internet is built on the IPv4 address model and is running out of addresses. Axel Pawlik, Managing Director of the RIPE NCC, the not-for-profit organisation that manages Internet addresses (not only assigning them, but mapping them and tracking usage) in EMEA and Central Asia says that we have just run out in North America, and that he thinks we are approaching a tipping-point for IPv6 adoption. There’s an IPv6 FAQ here and about 70% of RIPE NCC members have IPv6 addresses now, although not all are using them. For a background on the IPv4 address issue see here.

According to The International Telecommunication Union, 3.2 billion people will be online by the end of 2015. ARIN, which looks after North America, has just reached IPv4 exhaustion (so 4 out of the 5 regional Internet registries have now run out of addresses). Which means that adding another billion addresses, for connected Things, say, is looking rather difficult without IPv6.

Nevertheless, IPv4 is not quite dead, although there is very little hardware left that can’t cope with IPv6. There are many workarounds where extra addresses can be multiplexed onto one IPv4 address. There is a healthy market in IPv4 addresses (the UK government has just made some money for the UK taxpayer by selling some of its spares); the regulators quite like having fewer addresses to worry about (although some of the workarounds used to cope with the paucity of IPv4 addresses make things harder for the tracking of addresses for law enforcement); and there are even potential IPv4 backwaters (like Africa) which haven’t used up their IPv4 address allocation yet.

So, to reiterate, why should we all be interested in IPv6? Well:

  • We really are running out of IPv4 addresses. It’s not a hard stop but eventually the workarounds will snap – the Internet of Things may be the final straw.
  • IPv4 workarounds add complexity – never a good thing – and increase management overheads.
  • IPv4 addreses are becoming increasingly expensive, as their scarcity increases.
  • Apple has just made support for IPv6 mandatory for IOS9 apps that will appear in its appstore, which affects app developers.

It is perhaps unfortunate that just about the only concrete thing IPv6 brings is a bigger address space. Originally, Internet Protocol Security (IPsec) was developed for IPv6, where it was mandatory, but it was re-engineered for IPv4 (given that the Internet is still largely IPv4, this is good) and made optional in IPv6 (possibly, a retrograde step). So, improved security is no longer a driver for IPv6. Nevertheless, although the Internet was originally designed for free communication amongst academics (rather than commercial security), the IETF is now fully aware of the need for making Security a fundamental part of the Internet design process. Read this Internet Best Practice document “Pervasive Monitoring Is an Attack“, for example.

Back to IPv6, what this all means is that organisations can still get away with ignoring IPv6 for a time, but that this attitude will, increasingly, add unnecessary costs and overheads – it will be seen as poor governance. There is no need to panic and introduce IPv6 immediately (unless some business imperative demands it) but any organisation that wants to be in the forefront of innovation, especially with IoT, should put IPv6 very firmly on its roadmap – perhaps it can be made to share resources efficiently with another network upgrade/maintenance initiative.