[At-Large] [lac-discuss-en] They're out of IPv4 Addresses!
h.raiche at internode.on.net
Sun Sep 16 01:16:40 UTC 2012
There is a twist to the IPv4/v6 discussion - and it's about money (surprise!) Geoff Huston (our own internet pioneer/guru) asks why the carriers aren't migrating to v6 - and instead dealing with the address shortage (at least in the APNIC region) through NATing. As he explains, with NATing, the carriers gain information about what IP addresses are going where - information they can flog to advertisers. It's part of the money shift from the carriers in an analogue, circuit switched world where the carriers got the money, to an IP world where they are missing out and the revenue is going to the content providers. (the WCIT issue about accounting rates no longer providing the carriers with the money they used to collect is relevant here). So the carriers are making do with NATing instead of migrating to the v6 world where, once again, one IP address connects directly to another address and no NATting (and no collecting information about users) in between between (except in the usual circumstances of corporates, etc)
On 16/09/2012, at 6:11 AM, Carlton Samuels wrote:
> Hi Sala:
> You misinterpret my skepticism as denial. Wrong.
> I'm there with you on all the facts you quote. Yes too, on all the trends
> you finger. My skepticism is born of the big 'the sky is falling' hoopla.
> And an analysis that is predicated on immediate and/or imminent scarcity.
> [Scarcity as in demand driven!]
> For the last several years, I've studiously followed the ARIN policy
> discussions. Its only after you see the details and couple that to a few
> other facts that it jumps out at ya; the details just don't support the
> great alarm hurry-up-and-do-this. Look critically at some of the data in
> context of where access and demand are growing; Africa and Asia. Then
> have another think.
> No, we will not have a collapse anytime soon. No, I doubt if IPv4
> addresses will ever be totally exhausted. At least not if some of the
> policy positions I see being promoted in the ARIN region are implemented!
> Yes, if and when that time comes, if someone belly up to your favourite RIR
> and asks for a **bank* *of addresses, here's the likely response, "well,
> we ain't got that kind. But here's another kind works just as well if not
> better! Machine you have might even be prepped for it, see. But you might
> have to do a few little things to let her fly". That conversation will not
> take place with a mom and pop operation, trust me. End user networks are
> not provisioned as the popular myth says.
> Troll the lists and you will see this is one topic I've studiously stayed
> away from. Not because I don't understand the issues; I ran service
> provisioning businesses absolutely dependent on ready availability of IP
> addresses. It is because I remain underwhelmed by the hype.
> The facts tend to the inconvenient. To the hype, that is. So we are agreed
> to disagree, albeit agreeably. Nope, the hype on this is over the top.
> And for the life of me I can't figure out why. In the Y2K storm I knew
> what it was; there were folks hellbent on picking some pockets.
> Maybe its because I am at a point where I've learned to distrust all
> revealed wisdom and orthodoxy. So I might be the heretic here. Oh well,
> if salvation comes......
> - Carlton
> Carlton A Samuels
> Mobile: 876-818-1799
> *Strategy, Planning, Governance, Assessment & Turnaround*
> On Sat, Sep 15, 2012 at 2:21 PM, Salanieta T. Tamanikaiwaimaro <
> salanieta.tamanikaiwaimaro at gmail.com> wrote:
>> Snip the whole thing always brings me back to that Y2K jingle.
>> I would disagree and here's why.
>> As I write this email, the world's population according to the Population
>> Reference Bureau is 7,087,683,026 as at 6:18am Sunday morning on 16th
>> September, 2012. The World Mortality Report of 2011 rates (produced by the
>> Department of Economic Social Affairs by its Population division) shows
>> that over time the mortality rates have generally gone down although there
>> are still wide disparities in levels of mortality across regions. See:
>> *End Users*
>> Whilst there are 7 billion people on the planet, the World Internet
>> Statistics (IWS) suggest that there are 2,267,233,742 internet users as at
>> December, 31, 2011. See: http://www.internetworldstats.com/stats.htm.
>> These users distributed would be something like this:-
>> AFRALO region - 6.2%
>> APRALO region - 49%
>> EURALO region - 22.1%
>> LACRALO region - 10.4%
>> NARALO region - 12%
>> *Can current supply meet demand?*
>> As the push for Universal Service continues to grow as we have seen from
>> recent trends in ICT development, liberalisation of markets, competition,
>> it is generally expected that the end users will grow. There already is a
>> demand on Internet Address allocations based on consumption. The pull on
>> address spaces is also linked to "consumerism". The innovations in science
>> and technology have brought us smart phones, smart computers, smart
>> refrigerators, smart homes, smart cars means that there is a "level of
>> demand" coming from the globe on a "finite pool of resources". We see
>> trends where as Innovations increase eg. Applications (App) and a single
>> phone can have an average of 5 Apps. Internet Traffic is definitely growing
>> because of many reasons, the desire to communicate, content driven
>> applications being some of the reasons. Whether it is an entrepreneur away
>> on a business trip can with a few strokes be able to check procurement of
>> goods, what's in the cash register etc or a Surgeon performing remote tests
>> and/or surgery, one thing is certain, consumption of the Internet will
>> continue to grow.
>> So the issue becomes, can "Supply meet demand"? Maybe and only for a little
>> while. There are Network Address Translators (NATs) that can only do so
>> much for a little while but all it is really at the end of the day is
>> buying time and waiting for the inevitable which is the pool of IPv4
>> resources will run out.
>> *The growth in global demand to communicate implies the need for
>> For as long as the assumption is true that internet usage will grow there
>> will be a strain on the address allocation. There are variables that affect
>> demographics droughts, tsunamis, global food crisis, water scarcity,
>> climate change, migration, conflicts and wars that affect demographics but
>> a steward and in this case RIPE NCC will try to ensure that there is
>> sufficient preparation that existing resources are conserved within reason
>> and at the same time encourage transition.
>> The threat aside from running out of internet addresses on the IPv4 front
>> is the ability for Networks not to be able to communicate. The IPv4 Network
>> cannot communicate with the IPv6 and there are ways to address these where
>> network owners can elect to opt for which ever methods of transition suits
>> Fortunately, the good news is that to be able to account for the current
>> and future demands on address space allocations, the IPv6 address
>> allocations were designed to enable seamless communication. So the only
>> challenge now is organising IPv4 to IPv6 transition. The key word is a
>> transition. As for end users there are many things that we can do as
>> ordinary end users to help prepare for the transition but that is for
>> another time.
>> Whilst some may say, that it's just vendors trying to sell their wares. The
>> reality is that vendors will always try to sell their wares and you can
>> help keep them accountable by getting your region and network providers to
>> publish feedback on their wares See:
>> Best Regards,
>> lac-discuss-en mailing list
>> lac-discuss-en at atlarge-lists.icann.org
> At-Large mailing list
> At-Large at atlarge-lists.icann.org
> At-Large Official Site: http://atlarge.icann.org
More information about the At-Large