[ppml] Policy Proposal 2005-1: Provider-independent IPv6

[Marshall Eubanks]

Hello;

On Apr 28, 2006, at 10:16 AM, Stephen Sprunk wrote:

> Thus spake "Marshall Eubanks" <tme at multicasttech.com>
>> On Apr 27, 2006, at 5:45 PM, Jason Schiller (schiller at uu.net) wrote:
>>> So I don't think it is unreasonable to think on the agressive side
>>> that there might be fairly good adoption of IPv6 by 2013.  My
>>> projections show 1.3M routes by 2013... That means if I want to
>>> upgrade a router now, and it takes 2 years to certify and fully
>>> replace my existing network and I want to depreciate the cost
>>> of the router over 5 years, then I need to buy routers that
>>> support 2.3M routes.
>>
>> Isn't this a typo, shouldn't it be "1.3M routes" ? Otherwise, I don't
>> see where it comes from.
>>
>> I actually think that these numbers are not too far out of line; the
>> question is, is the time line out of line ?
>>
>> The US Small Business Administration
>>
>> http://www.sba.gov/advo/research/us_03ss.pdf
>>
>> estimates that there were 5,696,600 employer firms in 2003, 99.7
>> percent of which are small firms
>> (i.e., there are 17090 non small firms, with 500 or more employees).
>> Of the employer firms, there are 2,262,695 with 5 or more
>> employees, and 1,237,198 with 10 or more employees. I think that a
>> reasonable upper bound for the
>> potential number of ASNs in the US right now is the latter number.
>> (Since the number of firms is  not increasing exponentially as fast
>> as routes are, and since I am resolutely in back of the envelope
>> mode, I will assume that these numbers will be constant over the next
>> decade.)
>
> 2005-1 effectively requires 512 hosts (and 2 pipes) to get a PIv6  
> block, so the expected number of assignments under that policy  
> should be 17k.  It's also reasonable to expect most of those  
> assignments will be minimum-sized, so ISPs will be able to keep  
> them down to 1 slot/org by filtering anything longer than /48.
>
> I don't think anyone here is proposing, even as a joke, giving PI  
> space to orgs with 5-10 employees/hosts.

Cool, but that's not quite the point. I don't see PI space being  
driven down further than that, so I regard
that (actually, 10 or more employees) as an _upper bound_. Anything  
that makes it lower makes my numbers
safer.

This email was engendered from a lunchtime discussion with Vince  
Fuller and Dave Meyer and others at RIPE.
My feeling was that the arguments about scaling etc. are futile  
unless you know "scaling until what ?",
and the only way to do that is to try and estimate firm upper bounds  
with assumptions that can be argued.

(For example, if you say that I am off by a factor in the total  
number of entities with PI space, or that
my doubling times are off, or whatever, it's easy to recalculate  
these numbers.)

Regards
Marshall

>
>> So, as a small multi-homer, I do not see a need to support 1.3
>> million routes in 7 years, but I do see a need to support 600,000,
>> which may not be that different in practice. What I frankly do not
>> see in all of these numbers is a reason not to support 2005-1.
>
> Ditto, though I come at it from a different angle: no matter how  
> much one may complain about PIv6, PIv4 is an order of magnitude  
> worse, so as long as v4 is still natively routed and policies are  
> similar, the IPv6 routing table size is not going to drive router  
> upgrades -- v4 and VPNs will.
>
> If anything, the cost of constant upgrades to support the bloated  
> v4 table should be viewed as motivation to get customers moving to  
> IPv6.  As much as one may dislike PIv6, one route/ASN is a lot  
> cheaper to build routers for than today's nearly 10 routes/ASN.
>
> S
>
> Stephen Sprunk        "Stupid people surround themselves with smart
> CCIE #3723           people.  Smart people surround themselves with
> K5SSS         smart people who disagree with them."  --Aaron Sorkin