[183321] in North American Network Operators' Group
Re: Production-scale NAT64
daemon@ATHENA.MIT.EDU (Bjoern A. Zeeb)
Thu Aug 27 09:00:04 2015
X-Original-To: nanog@nanog.org
From: "Bjoern A. Zeeb" <bzeeb-lists@lists.zabbadoz.net>
In-Reply-To: <CAD6AjGTv0DCtmmdydmTDnFoZ7p2rU0HG4ai0JDew7ADkJAdhxw@mail.gmail.com>
Date: Thu, 27 Aug 2015 12:59:55 +0000
To: Ca By <cb.list6@gmail.com>
Cc: "nanog@nanog.org" <nanog@nanog.org>
Errors-To: nanog-bounces@nanog.org
> On 26 Aug 2015, at 15:23 , Ca By <cb.list6@gmail.com> wrote:
>=20
> On Wed, Aug 26, 2015 at 8:16 AM, <Valdis.Kletnieks@vt.edu> wrote:
>=20
>> On Wed, 26 Aug 2015 07:28:08 -0700, Ca By said:
>>=20
>>> Another relevant metric, less than 25% of my mobile subscribers =
traffic
>>> require NAT64 translating. 75+% of bits flows through end-to-end =
IPv6
>>> (thanks Google/Youtube, Facebook, Netflix, Yahoo, Linkedin and so on
>> ...).
>>=20
>> So I'm guessing that 75% of the traffic flows with better latency =
than
>> the 25% IPvhorse-n-buggy traffic? ;)
>>=20
>>=20
> Facebook says IPv6 is 20-40% faster
>=20
> =
http://www.internetsociety.org/deploy360/blog/2015/04/facebook-news-feeds-=
load-20-40-faster-over-ipv6/
>=20
> Another way to look at it, IPv4 is 20-40% slower than IPv6.
The question I have not seen the answer yet to is =E2=80=9Cwhy?=E2=80=9D
Is this really because of the network, e.g., separate pipes in some =
places still, with forwarding devices handling a lot less pps?
Is it because of people having done a newer cleaner-cut network stack =
implementation and lately cared about its performance?
Is it about middle nodes?
Has anyone done the research on this?=
