[6bone] IPv6 routing table size

[Michel Py]

Jeff,

>>> Bill Manning wrote:
>>> a single "million" is either way too large or
>>> way too small.  Depends on if you beleive the
>>> aggregation model will hold.

>> Michel Py wrote:
>> Agree.

> Jeff Thomas wrote:
> The bullet point from the presentation actually was
> "Routing table > 1M".

Are you referring to the one below that Jordi just posted?

> Jordi Palet wrote:
> Ok. Then you mean 863 ... the presentation is here 
> http://www.ipv6.net.cn/event/presentation/ipv6-china-report.pdf

If yes, I don't see where the scoop is. There are routers that handle
200k routes. Having people designing a router to hold 1M is yesterday's
news. Without specific info, I'd bet you a buck that Cisco and Juniper
are working on it, too. The fact that they are working on it does _not_
mean we will see a 1M prefixes routing table. A distinct possibility if
no multihoming solution is released soon, put still only a possibility.


> I guess the point of the aggregation model holding is the root of
> my question.  I was thinking it would hold, but it sounds like
> nobody is banking on it.
> What factors are considered here in holding the aggregation model
> in question?

The availability of a real IPv6 multihoming solution.

Michel.