[arin-ppml] Routing table growth, was Re: IPv4 is depleted today

[Tom Vest]

On Sep 3, 2008, at 10:42 AM, Iljitsch van Beijnum wrote:

> On 3 sep 2008, at 16:17, Howard, W. Lee wrote:
>
>> before the
>> inflation of the routing table presumed in this thread?
>
> I must have missed that... What was said about that?
>
> The only way the routing table is going to grow significantly faster
> than it has until now when we run out of v4 space is if the people
> that now get blocks of a million addresses are going to string
> together hundreds of small blocks. And they're not going to do that,
> because it's too much work and too expensive. If you're Comcast and
> already have a /8 (from ARIN--not a legacy one!) you can easily take a
> million addresses and NAT that space so you can put 2, 4, 8, 1000
> customers behind each of those addresses without going through complex
> and presumably expensive address acquisitions.
>
> The 90% of the address allocations for small blocks (10% of the space)
> will simply continue in the same way they have for years, someone who
> needs a /24 will get a /24 and not four /26s. Or maybe they'll get
> _one_ /26. This won't impact the routing table growth much, if at  
> all.
>
>>> In other words, IPv6 is not the only choice. The best choice
>>> maybe, but don't for an instant imagine that  properly
>>> motivated people can't find a way to make the IPv4 Internet
>>> continue to grow.
>
>> I do indeed imagine that.  It keeps me awake at night.
>
> You'll be able to go to http://www.google.com/ over IPv4 for a very
> long time. Don't expect do much peer-to-peer communication over IPv4
> in the less immediate future, though.
>
>>> That approaches the real question: IPv6 migration has a cost
>>> function and so does IPv4 growth. Where do the cost lines
>>> cross? When does it become less expensive to deploy IPv6 and
>>> exert the push that brings
>>> IPv4 to a close?
>
>>> The plain truth is that we don't have enough data to do
>>> better than guess at the answer to that question.
>
>> I agree with your question.  What data would we need?
>
> http://www.networkworld.com/news/2003/1215ipv6.html
>
> (That was 5 years ago, BTW.)
>
>>> But we do
>>> know that each IPv6 route costs 2 to 3 times as much to
>>> implement in the router as an IPv4 route.
>
> Do we? More bits = more expensive, presumably. But where does 2 to 3
> times come from? Note that today we have about 8.4 prefixes per AS for
> v4 and 1.2 for v6.
>
>>> And we know that there is no new magic in IPv6 routing:
>>> multihoming still requires a unique global route.
>
> There is an alternative that is especially appropriate for smaller
> multihomers:
>
> http://www.ietf.org/html.charters/shim6-charter.html
>
>>> If I had to guess, I'd guess that the cost lines never cross. Even  
>>> as
>>> IPv4 gets more expensive, it remains more cost effective to
>>> squeeze a little more out of it than to attempt an IPv6
>>> migration.
>
> It's not the cost: it's the benefits. If it costs $50 to implement v4
> and you can reach 500 sites out of the Alexa top 500, that's $0.10 per
> destination. If v6 is only $40 but you can only reach the 0.4% of the
> Alexa top 500 sites that have IPv6 that's $5 per destination.

The question is, what mechanisms will contribute to increasing that  
benefit number?

1. Incumbent IPv4-based operators altruistically make their own  
content and services transparently accessible to remote IPv6-only  
networks,

and/or

2. Topologically adjacent IPv6-only networks grow and proliferate,  
driven solely by IPv6-based demand for IPv6-based content and services,

and/or

3. IPv4 transfer prices and/or outsourced v4/v6 translation services  
remain low enough so that otherwise IPv6-only networks can exchange  
traffic with each other remotely without being substantially /  
differentially disadvantaged relative to IPv4-based incumbents.

Is the community skeptical about (1)? We seem to be quite cynical  
about other presumptions about altruistic behavior, or even neutral/ 
status quo/policy compliant behavior in the address scramble to come.  
(1) would not only cost operators $$$ with no obvious/direct return,  
it would also tend to moderate the demand/prices for IPv4 post-runout,  
which would not be in the interests of potential IPv4 sellers.

As assumptions about (2) realistic, especially in the presence of  
ongoing, active competition by full-functioning IPv4-based operators?

It seems like an awful lot is riding on faith in (3), even though it  
directly conflicts with (1).

TV