[189570] in North American Network Operators' Group
Re: IPv6 is better than ipv4
daemon@ATHENA.MIT.EDU (Christopher Morrow)
Thu Jun 2 13:07:27 2016
X-Original-To: nanog@nanog.org
In-Reply-To: <CAD6AjGQGyaPAaOJCTpfDsd2qLWtnjtr0Lv4X2_7Q__Y=1bRytQ@mail.gmail.com>
Date: Thu, 2 Jun 2016 13:07:23 -0400
From: Christopher Morrow <morrowc.lists@gmail.com>
To: Ca By <cb.list6@gmail.com>
Cc: "nanog@nanog.org" <nanog@nanog.org>
Errors-To: nanog-bounces@nanog.org
On Thu, Jun 2, 2016 at 12:55 PM, Ca By <cb.list6@gmail.com> wrote:
>
>
> On Thursday, June 2, 2016, Christopher Morrow <morrowc.lists@gmail.com>
> wrote:
>
>>
>> On Thu, Jun 2, 2016 at 12:23 PM, Daniel Corbe <dcorbe@hammerfiber.com>
>> wrote:
>>
>>> Maybe we should let people believe that IPv6 is faster than IPv4 even i=
f
>>> objectively that isn=E2=80=99t true. Perhaps that will help speed alon=
g the
>>> adoption process.
>>
>>
>> =E2=80=8Bdo we REALLY think it's still just /marketing problem/ that kee=
ps v6
>> deployment on the slow-boat?=E2=80=8B
>>
>>
> YMMV, but the majority of my customers are ipv6. And for those customers
> with ipv6, 73% of their traffic is e2e IPv6.
>
>
=E2=80=8BI understand that tmo's (us at least) network is v6 native to the
handset... my question was really trying to point out that even if tmo is
100M customers, there are ~3x that on sprint/vz/att/etc ... so just in the
US is 25% repreesntative? and outside the US what does the mobile address
family spread look like?=E2=80=8B
then, what if the resource being accessed to by the mobile users in
zimbabwe are local to zimbabwe and there's only ipv4 versions of that
mobile service... the reasons to go v6 on both sides aren't as clear. (to
me)
> I agree that there are many dark corners of Santa Cruz without IPv6, but
> the story is: the whales of content and eyeballs are on IPv6, and it is
> cheaper (no cgn) and faster (RUM data) than the ipv4 alternative.
>
>
=E2=80=8BI get why things look better in the cases of FB/TMO (as one exampl=
e)...
but selling 'you should ipv6 because FB/TMO is better!' =E2=80=8Bisn't real=
ly true
all ways.
> Does it really matter what single digit % of Alexa 1M has a AAAA?
>
=E2=80=8B:) I have no idea...=E2=80=8B
