Waste will kill ipv6 too

Lee Howard lee at asgard.org
Wed Dec 20 19:16:24 CST 2017

On 12/20/17, 1:23 PM, "NANOG on behalf of Mike" <nanog-bounces at nanog.org
on behalf of mike-nanog at tiedyenetworks.com> wrote:

>On 12/17/2017 08:31 PM, Eric Kuhnke wrote:
>> some fun examples of the size of ipv6:
>> https://samsclass.info/ipv6/exhaustion-2016.htm
>Every time I see these "Look how many more addresses we have now with
>IPv6", I just shake my head.
>  Yes, the address space is very large. But, all of the protocols, all
>of the addressing guides, all of the operational 'best practices', ALL
>OF IT, increases by orders of magnitude the amount of waste along with
>it. Call this the 'shavings', in IPv4 for example, when you assign a P2P
>link with a /30, you are using 2 and wasting 2 addresses. But in IPv6,
>due to ping-pong and just so many technical manuals and other advices,
>you are told to "just use a /64' for your point to points. So, the
>actual waste is dilutes the actual implementable size of the total ipv6
>address space due to the waste component. And I have not yet seen any
>study or even proposed theory to explore what the IPv6 Internet would
>look like, if used in place of all IPv4 in all the places and ways that

Okay, so do the thought experiment.

Let’s say 10 billion people in the world.
Let’s say each of them has 10 devices, each with a /64, and to be
ridiculous, each with a p2p link to the others, so we have 100 /64 per
That’s 1 trillion /64s.
Oh, dear, I’ve used up the first /27 of IPv6 space.

What if we try giving ten /48s to each of 10 billion people, one for each
of the ten providers they’ll have? Sure, I’m handwaving the p2p links, but
that’s why we assign that /48.
That’s 100 billion /48s, which is something like a /11.

I’ve tried several times to come up with a scenario that leads to
depletion in less than 200 years, and I haven’t managed it. Can you do it?

> why is
>nobody thinking or talking about the looming exhaustion of ieee OUI

> Network cards made 15 years ago and since consigned to the
>electronics scrap heap in the sky, take with them their addresses never
>to be reused again (unless you are a freak like me and keep some for
>'positively never assigned anywhere'). And old dead companies that were
>assigned OUIs, they get 24 bits of address space to take to their
>graves. We should be re-thinking mac addressing altogether too.

Like EUI-64?  https://standards.ieee.org/develop/regauth/tut/eui.pdf


More information about the NANOG mailing list