Data Centers

IPv6 and the Internet of Things

Nick Hardiman explains why the switch to IPv6 will further enable the world of connected devices.

The Internet of Things means the continued growth and development of an open, global Internet. The tool that is going to make that possible is IPv6, and organizations from the IEEE to ISOC have been pushing it (World IPv6 Launch Day was June 6, 2012). But what is it good for? Is it just about replacing IPv4 addresses like "" with weird looking IPv6 addresses like "2001:41c8:51:27e::10"?

Getting from nothing to something

IPv4 was the main tool used to create the global Internet. It was an IT success story, of getting from nothing to something for everyone. It wasn't perfect, but the history of IT success stories is full of examples of dodgy decision-making. Using two ports for FTP probably seemed like a good idea at the time. Tying Ethernet MAC addresses to their manufacturers probably wasn't the greatest security decision. A URL really doesn't need :// stuck in it.

As an example, take the IPv4 address. It looks something like this: That IPv4 address format is called dot decimal notation, and it is a little whacky. It's a coding system that turns the 32 bits that a machine can read into four decimal octets for humans to read. It isn't really decimal, it isn't really binary, and it certainly isn't octal. This dot decimal system helped IT people remember IP addresses, at a time in IT history after they had agreed that splitting code into 8 bit chunks was useful and before they had realized this shorthand notation would escape and find its way into every home and office in the world.

Getting from something to something better

Jumping from version 4 to version 6 will simplify the networking world. The simple fact that IPv6 has a zillion addresses means the workarounds used with IPv4 can be removed. The simple Internet model of sending a packet from one address to another was broken with NAT (Network Address Translation).

With IPv6, there is no need for private addresses. Most enterprise computers have private IP addresses, so making the most of IPv6 in an enterprise will require quite a shift in working practices.

IPv6 can also make current practices redundant -- IP address block land grabs, the IP pools of ISPs, and even the client/server model could disappear.

Should I switch to IPv6 in 2013?

For an enterprise, yes - well, partly. For a residential user, no.

When an enterprise refreshes its technology, it makes sense to have a strategy that insists on new kit that can handle IPv6. It makes sense to build new IPv6  services alongside new IPv4 services. There is no pressure to upgrade existing services to use IPv6.

At home, most of the components needed to make sure IPv6 replaces IPv4 are already in place. The popular operating systems can talk IPv6. Most home computers understand IPv6, and so do many home routers. Many Internet services have IPv6 addresses. Unfortunately, many people can't reach them because their ISP only does IPv4 - if a residential early adopter wants to reach an IPv6-only website, she may be out of luck. It's going to be a while before IPv6 becomes the norm in domestic networks.

Can I face the change?

And there's the learning curve that comes bundled with every new technology. I know I ought to learn about IPv6 and change my servers that only use IPv4, but I am not excited by the prospect. It feels like a low-priority chore, and keeps slipping to the bottom of my to-do list. After many years in the Internet industry, I remember IPv4 addresses like I remember birthdays (the really important ones stick, but I need reminding about the rest).  I haven't thrown enough IPv6 addresses at my memory for any to stick.

I will learn, and I will change my servers. IPv4 enabled the Internet and IPv6 will enable the Internet of Things. If I want a future in Internet IT, I must update my skills. IPv6 and the Internet of Things are the future.


Nick Hardiman builds and maintains the infrastructure required to run Internet services. Nick deals with the lower layers of the Internet - the machines, networks, operating systems, and applications. Nick's job stops there, and he hands over to the ...


Every local network has it's own Internet Address on it's external router, but everything inside that network is isolated from the internet outside. This means that IPV4 can easily handle any number of devices internal to your company/home network. The fact that most external addresses to homes and many businesses are dynamic further complicates this matter, meaning that communication between devices in different networks need to use the DDNS protocol. External router addresses and DDNS can easily handle IPV4 and/or IPV6. This has no effect on your internal network. Now unless you have more than 4G network enabled devices on your network, then you will not need IPV6. IPV4 is faster and can handle any number of devices that you are ever likly to encounter in one company/home. Communication between devices in one private network and another never really happens unless through a routing protocol like DDNS or something like that. i.e. Those protocols are already in place and can handle IPV4 local networks just fine.

i.e. My conclusion is that IPV4 is all that is every likely to be needed by networked devices for which the IoT buzzword was invented.

i.e. This article does not address the practical need for IPv6 (it is not needed for IoT).

Lawrence Garvin
Lawrence Garvin

An interesting take on IPv6, although my expectation is that outside of mobile/wireless service providers (Verizon, AT&T, Sprint, T-Mobile), there’s not likely to be a lot of IPv6 adoption.. ever. There’s no compelling reason for a private organization to go through the growing pains of implementing IPv6, or abandoning IPv4. Existing IPv4 private address space is more than sufficient for almost any organization. Furthermore, for those organizations that still have legacy systems deployed (WinXP/2003), IPv6 is not even an option. Ironically, while you've said “no IPv6” for residential users??? which I don't necessarily disagree with because of the complexity factor, the irony is that HOME networks are probably the most likely small networks to take advantage of IPv6 – because it is fully automatic and won’t require IPv4 DHCP to do peer-to-peer networking in a home environment. But, as noted, the ISP's aren't doing IPv6 either -- because they don't need it. Then again, they don't need to -- all they need is a router that does NAT-PT (


Let me get this straight. 'Internet of Things' ( IoT) is in reality, a euphemism for 'Internet of Stuff'. Now, if this follows, then the reality of IoS is iOS which means that Apple wins. Right?

Editor's Picks