IPv6 Is Here, Get Used to It

Geoff Huston published an interesting number-crunching exercise in his latest IPv6-focused blog post: 8% of the value of the global Internet (GDP-adjusted number of eyeballs) is already on IPv6, and a third of the top-30 providers (which control 43% of the Internet value) have deployed large-scale IPv6.

The message is clear: The big players have moved on. Who cares about the long tail?

Obviously this message won’t persuade the naysayers who still believe IPv6 will never happen. They must be too young to remember what happened in the early 1990’s when some people started deploying something called IPv4.

At that time we were happily living in crazy worlds of DECnet, SNA, IPX, AppleTalk or whatever other acronym your preferred vendor supported, and there were those crazy people talking about IP. Nobody really believed IP would take on… until it did. It took years (every transition takes longer than expected), and it was more sudden than anyone believed (see also: punctuated equilibrium). We might be pretty close to that point with IPv6.

On a somewhat tangential note, I had the dubious privilege of being able to listen to a roundtable at (otherwise excellent) recent Slovenian IPv6 summit. The roundtable featured several local operators practicing group lamentation and yammering. Supposedly majority of them already deployed IPv6 in their networks and services, but almost nobody would enable it by default for residential customers due to lack of business case and perceived lack of content on IPv6.

Hint: You don’t need a business case for IPv6. It’s a business continuity solution.

Unfortunately not one of them realized that the rest of the world doesn’t care. The big guys (connectivity and content providers alike) have moved on and will build the next-generation Internet regardless of the denial of everyone else.


  1. I can attest to that in a major globally known retail company. Management are of the belief that they can wait for 'others' to change, then they'll follow. IPv6 isn't even thought about or considered, the eCommerece mgmt team have no idea what IPv6. In the meantime management have said that we will be utilising the various transition mechanisms that surely will be put out there by providers to ensure that nobody is 'really' cut off, despite them getting a poor quality experience. It drives me mad!!
  2. "You don’t need a business case for IPv6. It’s a business continuity solution."

    Couldn't agree more!

    When developing our IPv6 consulting collateral 4 years ago, this was (and still is) my answer to the customer objection - "but where's the business case". You plan for floods in your comms room, tornadoes hitting your midwest data center, but you have no plans for something far more certain than an "act of God"?
  3. There's nothing that prevents the Internet from relying solely on IPv4 for the next decade.

    IPv4 blocks are not actually scarce. Sure the RIRs are out of free IPv4 blocks, but there's still large quantities of unused IPv4. Is IPv4 space expensive? Possibly. So far it's cheaper than rolling out IPv6.

    Home internet users will get stuffed behind NAT and service providers will reallocate their IP block either to their own cloud services, or will hold their IPv4 allocations as ever-appreciating assets.

    A cloud service provider can charge $16/IP per month with IPv4. With IPv6, in theory, that income stream would go away. There's no end-to-end support for IPv6 yet in EC2, Azure, VMWare vCloud Air.

    I expect to see a contraction of IPv4 usage over the next decade as more business opt to run their internet facing data centers in "the cloud."

    I love IPv6. I have been trying to find a company that would let me implement IPv6. But the business case of "business continuity" falls flat. No one is out of IPv4 space. It just costs money to get more IPs now.
    1. "There's nothing that prevents the Internet from relying solely on IPv4 for the next decade."

      You must be living in a nice bubble. I can tell you stories about cloud services not being launched because they SP could not get IP addresses, SP nightmares due to NAT-related logging (required by the law), other SP nightmares because they cannot use NAT logging (triggered by privacy concerns) and yet cannot get new IP addresses for their customers, enterprises becoming Internet Registries (LIRs) just to get a /22...

      Obviously nothing along these lines would ever persuade people living in total denial. In the meantime, as I wrote, the big players have moved on.
  4. IPv6 is must, The most obvious answer is that IPv4 is out of IP addresses. IPv4 has only 4.3 billion addresses, and with PCs, smartphones, tablets, gaming systems, and just about everything else connecting to the Internet we've tapped the system dry. IPv6 uses 128-bit addresses and is capable of 340 undecillion addresses.
    1. IPV4 is not out of addresses.
      My company has a /16. We have among 2 groups .. maybe 70 or so entries in our DNS. If other folks were even 1% of the usage rate we do that would be 200+ Billion devices and not even many of those need to talk to anything outside their local network.

      Not a single non-DMZ natted device on my networ needs a unique Internet addressable address so why worry about giving them one.

      I own a /20 from the 90s I am not using - I am sure that there are many companies and individuals like me.

      We're not out of IPV4 - we are out of easy IPV4.
    2. Being out of easy IPv4 is effectively being out of IPv4. Even if all the hard IPv4 could be recovered it wouldn't be enough for future demands.

      There are sufficient drivers for IPv6 to eventually have all ISPs and Mobile operators providing it. Content providers will deploy it too to optimise their site performance. Eventually enterprise networks will get on board but by then it will already be inevitable.
Add comment