ICANN begins migration to the IPv6 protocol

The Internet Corporation for Assigned Names and Numbers—the company in charge of managing IP addresses and domain names—has kicked off its transition to the IPv6 protocol, according to CNet. ICANN says it has assigned IPv6 addresses to six of the world’s 13 root server networks, which regroup databases that reference all top-level domain names. Internet service providers have welcomed the development “as part of the continuing evolution of the public Internet.”

The switch to IPv6 may not happen overnight. CNet quotes Jay Daley, director of IT for Nominet (the company that handles UK domain names), as saying, “If you really want to see take-up of IPv6, we need the people who run high-volume Web sites to switch over to providing both IPv6 and IPv4 access to them. There are very few sites out there that do that.” However, the transition should be accelerated by the increasing scarceness of IPv4 addresses. According to Daley, “It may soon be extremely difficult to get hold of any more IPv4 addresses.”

The IPv4 protocol uses 32-bit addressing, which limits the total number of possible addresses to 232, or 4.29 billion. By contrast, IPv6 uses 128-bit addressing, thus enabling a virtually inexhaustible supply of addresses (2128, or 3.4 × 1038 in total). IPv6 addresses are usually presented as eight groups of four hexadecimal digits, for example 2001:0db8:85a3:08d3:1319:8a2e:0370:7344. Luckily for administrators of local networks, IPv6 addresses with adjoined groups of zeroes can be shortened from, say, 2001:0db8:0000:0000:0000:0000:1428:57ab to just 2001:0db8::1428:57ab.

Comments closed
    • leeroy
    • 12 years ago

    Could somone clear up a quesion for me 🙂

    Why would i need to upgrade my INTERNAL network to v6 when it works very well on v4. I have more than enough addresses available on different subnets for different offices all VPN’d together.

    All i should have to do is upgrade my network gateways to v6 so that the external interfaces have a nice long and cumbersome unique v6 IP. Ensure the gateway device supports some new fangled IPv6 to v4 translation and use v4 addresses on my internal network.

    I actually like the added security of NAT, would IPv6 not remove this barrier and make my network devices more susceptible to attack ?

    Am i missing something really obvious ?

    • albundy
    • 12 years ago

    so, what your saying is that we will have better quality pr0n, right?

      • ludi
      • 12 years ago

      Uh…if 2001:0db8:85a3:08d3:1319:8a2e:0370:7344 is the kind of thing that really floats your, ehm, boat, then sure. Lust away.

    • ssidbroadcast
    • 12 years ago

    Man, that thing looks like a password for a crappy NES game.

      • eitje
      • 12 years ago

      DON’T YOU DARE TALK DOWN ABOUT METROID.

    • provoko
    • 12 years ago

    I would shoot myself if I had to use the address 2001:0db8:85a3:08d3:1319:8a2e:0370:7344. Haha.

      • eitje
      • 12 years ago

      you mean you would PING yourself!

        • eitje
        • 12 years ago

        woah. actually, there’d be enough IP addresses at that point to be able to assign them out to people when they’re born, like SSNs in the US.

        imagine THAT.

          • Krogoth
          • 12 years ago

          Acutally there enough addresses to label all living macroorganisms with an unique address on the planet!

          • SuperSpy
          • 12 years ago

          Taken from the IPv6 article on ars technica:

          l[<340,282,366,920,938,463,463,374,607,431,768,211,456

          To put this into perspective: there are currently 130 million people born each year. If this number of births remains the same until the sun goes dark in 5 billion years, and all of these people live to be 72 years old, they can all have 53 times the address space of the IPv4 Internet for every second of their lives. Let nobody accuse the IETF of being frugal this time around.<]l

            • eitje
            • 12 years ago

            well, let’s hope that in those 5 billion years, immortality isn’t discovered!

            • Krogoth
            • 12 years ago

            Well, immortality is an impossiblity in the absolute sense.

            You have to defy laws of thermaldyanmics in order to obtain it. 😉

            • titan
            • 12 years ago

            Oh yeah? Then just how do you explain Duncan McCloud?

          • Flying Fox
          • 12 years ago

          I remember hearing from somewhere that 2^128 is in theory more than the number of atoms in the universe or something?

            • Viathon
            • 12 years ago

            I’ll be sure to let you know… still counting…

    • UberGerbil
    • 12 years ago

    It’s an ugly, sordid tale. IIRC, the IETF wanted to support IPv4 addresses within the larger IPv6 address space by allowing a notation that looks like
    X:X:X:X.d.d.d.d (such as FF::FF.65.128.1.1) They had to use a different delimiter so they could distinguish this case from one where you had IPv4 digits decimal digits that could be interpreted as hex, for example FF::FF.65.40.1.1
    Such an address just using dots — FF..FF.65.40.1.1 — would be ambiguous because you wouldn’t know whether to interpret that 3rd to last number as decimal 40 or as hex 40 (decimal 64) for example.

    So they went with colons. Unfortunately, this notation creates other problems because of the use of the colon in port addresses, ie §[< http://69.65.116.162:80<]§ Consequently, they had to invent /[

      • titan
      • 12 years ago

      As far as I know, the IPv4 in IPv6 is in the format of “::192.168.0.0” Note the leading double colon. I could be wrong though.

      IPv6 has been pretty tough for me to grasp some concepts, but I think it’s the material I’ve been reading that’s been giving me trouble. Part of the problem is that I just drive the car and I don’t know what’s under the hood.

    • Krogoth
    • 12 years ago

    It is only happening in EU, and other nations that their internet infrastructure is relatively young and growing.

    IPv4 is going to around in the USA for a long, long time. It will end-up being like mertic versus USA imperial measurement systems.

      • indeego
      • 12 years ago

      Except not. Eventually everyone will use v6, because they will have to for network expension, and then they will abandon v4.
      A better analogy is the phone system adding on area codes. And more codes, and more.

      This is quiet unlike metric/imperialg{<.<}g

        • Krogoth
        • 12 years ago

        You underestimate the infrastructural problem.

        USA has invested tons of its networking infrastructure with IPv4.

        It is a more involved than changing protocols and do firmware updates.

        A lot of preexisting networking equipment in USA simply lacks the memory and processing power to handle heavy IPv6 traffic. The cost of upgrading does not yield the returns desired by majority of businesses.

        I give IPv6 widespread adoption in USA at least 20 years if not more.

        There is a reason why IPv6 has legacy support for IPv4. 😉

          • Usacomp2k3
          • 12 years ago

          May I ask where you get this information from? I would think that the USA would have more motivation so upgrade because of it’s widespread use, and also the capital to do so.

            • Krogoth
            • 12 years ago

            Think about it.

            Remember, how resistent are companies to upgrades they are not really necessary? IPv6 is fine and dandy. The real question to any company, is it worth the hassle, loss of reveune, time to do the upgrade?

            That is if anything the #1 barrier for widespread IPv6 adoption in the USA.

            The proverb “If ain’t broken don’t fix it”. Rings ever more true.

            • titan
            • 12 years ago

            If the US is going to continue communicating with the rest of the world, it must switch over to IPv6. Also, IPv4 is broken. Reportedly, there are a lot of things that are easier to do in IPv6 than IPv4, and more secure as well. Support for IPv4 will disappear someday.

            In short, the IPv6 is necessary, so it will happen sooner rather than later.

            • Krogoth
            • 12 years ago

            That is why IPv6 spec has build-in IPv4 legecy support via tunneling. 😉

            • TheTechReporter
            • 12 years ago

            The point is that the IPv4 legacy support (and other features) will make it painless to migrate from v4 to v6.
            In other words, the “hassle, loss of reveune, time to do the upgrade” _won’t_ add up to much, so companies will make the change relatively quickly.

            Also, I have to laugh about people complaining about the new IP address format. In 2008, how often do you expect to _manually_ type in an _IP address_? “Never” comes to mind. Talk about a moot point.

            • Krogoth
            • 12 years ago

            Ummmm… you have manually type-in addresses for some servers and for references with name resolution.

            • indeego
            • 12 years ago

            End users don’t though. Techs just deal with it.

            And the comment about security being better on IPv6 is an unproven one. Yes there is tech to improve security, but there is also a chance that something untested exists and hasn’t been exploited yetg{<.<}g

          • titan
          • 12 years ago

          Exactly right, except it won’t take the time you’re quoting. I’m guessing two to three years before we start seeing IPv6 addresses being used instead of IPv4 on a regular basis.

          Part of the beauty of IPv6 is that it can be IPv6 equipment can be implemented anywhere in the network without support being need above or below it in the chain. ISPs can take a top down approach or a ground up approach. Either way they go, I’m sure they’ll grandfather the new tech in as they do now.

    • Usacomp2k3
    • 12 years ago

    Maybe Google will switch. And microsoft/yahoo.

    EDIT: is /. on ipv6 yet 😉

    • flip-mode
    • 12 years ago

    Why change from a period to a colon?

      • willyolio
      • 12 years ago

      because colons have TWO dots for DOUBLE the… um… internettability.

      • Meadows
      • 12 years ago

      They have to.

      I’m studying IT at school and we went through this topic as well during networking classes. As much as they didn’t expect it, the world is running out of IP addresses (as per the v4 standard). It doesn’t matter how they try segmenting the networks or attempt adding tricks, they’re simply running out. IPv6 has been waiting to jump in for a long time now.

      The situation is quite similar to the 32-bit and 64-bit operating system RAM thing. Back when the first 32-bit systems appeared, people thought nobody will ever reach the top and/or need even more. Then they did, but thankfully 64-bit systems are such a leap that they’ll provide headroom for decades from now. The same happened to IP addresses, they need to extend the headroom because soon people can’t get addresses anymore.

        • Meadows
        • 12 years ago

        I don’t think this was meant as a reply here, but nevermind. 😉
        As for the colon thing, it’s better differentiation. I guess it’s an IT standard too or the like.
        Think of MAC addresses. They are also hexadecimal and use colons as separators.

      • titan
      • 12 years ago

      IPv6 is written as a hexadecimal representation instead of decimal. I can’t remember where I read it, but writing it this way makes it 75% shorter. Or something.

        • evermore
        • 12 years ago

        Hex reduces 4 binary digits to 1 hex digit, so it’s 75% smaller than binary.

        The reduction from dotted decimal will vary and may even be an increase depending on notation. 4.2.2.2 in dotted decimal is 4:2:2:2 in hex (or could be 04:02:02:02), but 255.255.255.255 is FF:FF:FF:FF. The maximum reduction possible would be 50% from dotted decimal to hex, as 15.15.15.15 in hex without zeroes is F:F:F:F. Anything higher than 15 would result in 2 significant hex digits, and of course the majority of numbers are higher than 15 (infinitely higher theoretically, but up to 240 higher in 8-bit binary).

        At any rate, this is going to be a pain in the ass. No matter what reduction techniques there are, many people, whether in a technical support or administrative capacity or just a knowledgeable user, often find need to resort to IP addresses for testing or for simply being easier than figuring out what it is in DNS. I know what 4.2.2.2 is, I couldn’t possibly remember what it resolves to in DNS (granted 4.2.2.2 is a simple IP but I also know what 216.126.50.42 is on my company network and can never remember the entire FQDN). After all, the majority of IPs won’t actually have those strings of insignificant zeroes (and you can’t even reduce multiple blocks of them if they aren’t a single string).

        I’d like to smack around whoever thought “a string of 32 digits, yeah that’ll be readable and memorable, put it in the manual”.

Pin It on Pinterest

Share This