[141443] in North American Network Operators' Group
Re: Microsoft's participation in World IPv6 day
daemon@ATHENA.MIT.EDU (Owen DeLong)
Wed Jun 8 09:35:47 2011
From: Owen DeLong <owen@delong.com>
In-Reply-To: <BANLkTikhqEwr9383UFQ7h+3kL2WX_CvXtyc1_GwNCWAVsh+eYA@mail.gmail.com>
Date: Wed, 8 Jun 2011 06:30:55 -0700
To: Cameron Byrne <cb.list6@gmail.com>
Cc: nanog@nanog.org
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org
On Jun 8, 2011, at 6:09 AM, Cameron Byrne wrote:
> On Wed, Jun 8, 2011 at 6:04 AM, Owen DeLong <owen@delong.com> wrote:
>>=20
>> On Jun 8, 2011, at 5:47 AM, Cameron Byrne wrote:
>>=20
>>> On Wed, Jun 8, 2011 at 12:09 AM, Owen DeLong <owen@delong.com> =
wrote:
>>>>=20
>>>> On Jun 7, 2011, at 9:59 PM, Martin Millnert wrote:
>>>>=20
>>>>> Owen,
>>>>>=20
>>>>> On Tue, Jun 7, 2011 at 11:47 PM, Owen DeLong <owen@delong.com> =
wrote:
>>>>>> LSN is required when access providers come across the following =
two
>>>>>> combined constraints:
>>>>>>=20
>>>>>> 1. No more IPv4 addresses to give to customers.
>>>>>> 2. No ability to deploy those customers on IPv6.
>>>>>=20
>>>>> 2 has little bearing on need of LSN to access v4. Insufficient =
amount
>>>>> of IPv4 addresses =3D> LSN required.
>>>>>=20
>>>>> Regards,
>>>>> Martin
>>>>=20
>>>> No, if you have the option of deploying the customers on IPv6, you =
don't
>>>> need LSN.
>>>>=20
>>>> The problem is that until the vast majority of content is =
dual-stack, you can't
>>>> deploy customers on IPv6 without IPv4.
>>>>=20
>>>>=20
>>>=20
>>> cough cough NAT64/DNS64 ...
>>=20
>> Doesn't solve the problem unless your users are all on cell-phone =
browsers
>> that don't do a lot of the things most users do with real internet =
connections.
>>=20
>=20
> Most of my users are on cell phone browsers :)
>=20
> Furthermore, i can choose which ones get ipv4-only NAT44 and which get
> ipv6-only + NAT64
>=20
> Now, only if there was major cell phone OEM support ....
>=20
>=20
> Also, i would like to extend the idea that as IPv6 becomes dominant in
> the next few years (pending access networks), the need for IPv4 access
> will wane and LSN for the IPv4 will become more acceptable as IPv4 is
> just the long tail.
>=20
Agreed... However, where I differ is that I believe it is content and =
services
which will drive the ability for IPv4 to be considered long tail. If all =
of the
content and services were IPv6-capable today, the need for LSN would
be very near zero (limited to the consumer devices that need to be
upgraded/replaced to understand IPv6.)
However, as it stands currently, a consumer would not consider an IPv6
connection with NAT64 or other LSN to be equivalent to what they expect
today (unless they're on a cell-phone where they already expect the
internet experience to be completely degraded).
Owen
