[arin-ppml] Opposed to 2010-9 and 2010-12

[Owen DeLong]

On Oct 10, 2010, at 9:07 AM, Kris Foster wrote:

> 
> On Oct 10, 2010, at 5:42 AM, Owen DeLong wrote:
> 
>> 
>> On Oct 10, 2010, at 5:37 AM, Scott Leibrand wrote:
>> 
>>> 
>>> On Oct 10, 2010, at 8:27 AM, Owen DeLong <owen at delong.com> wrote:
>>> 
>>>> 
>>>> On Oct 10, 2010, at 1:20 AM, Joe Maimon wrote:
>>>> 
>>>>> 
>>>>> Is there any reason a large ISP might want more than one 6rd scheme?
>>>>> 
>>>> Not that I can think of, and, the real problem with 6rd is that it doesn't matter what
>>>> size ISP you are... From the tiniest to the largest, they all consume the same
>>>> messed up prefix size for the same number of end bits to the end site.
>>>> 
>>>> So even a tiny ISP with 100 customers, if they want to give /56s to their end sites
>>>> will still consume 56-32 = /24.
>>> 
>>> Well, if you only have one v4 prefix, you don't need 32 bits.  That only helps in the smallest cases though. 
>>> 
>>> -Scott
>> 
>> While I haven't done the crawl through the routing table to be sure, I would posit that the number
>> of ASNs advertising a single prefix is probably less than 1%.
>> 
>> If someone with more time on their hands cares to validate this, it would be an interesting number
>> to have.
> 
> Total ASes present in the Internet Routing Table:                 34937
> Origin ASes announcing only one prefix:                           14711
> 
> (Philip Smith seems to email his routing table analysis everywhere except here)
> 
Sorry... I should have been more clear...

I would posit that the number of ISP ASNs advertising a single prefix is probably less than 1%.

The numbers above reflect ALL ASNs, not ISPs alone.

Owen

> --
> kris