[188212] in North American Network Operators' Group
Re: Cogent - Google - HE Fun
daemon@ATHENA.MIT.EDU (Owen DeLong)
Fri Mar 11 14:37:22 2016
X-Original-To: nanog@nanog.org
From: Owen DeLong <owen@delong.com>
In-Reply-To: <CAP-guGUqXWvgKNckGODXwLWo9fZfQSSafzawVhBFOefWqbnHBA@mail.gmail.com>
Date: Fri, 11 Mar 2016 11:34:36 -0800
To: William Herrin <bill@herrin.us>
Cc: North American Network Operators' Group <nanog@nanog.org>
Errors-To: nanog-bounces@nanog.org
> On Mar 11, 2016, at 06:16 , William Herrin <bill@herrin.us> wrote:
>=20
> On Fri, Mar 11, 2016 at 7:40 AM, Jon Lewis <jlewis@lewis.org> wrote:
>> On Thu, 10 Mar 2016, William Herrin wrote:
>>> It's Cogent's fault because: double-billing. Google should not have =
to
>>> pay Cogent for a service which you have already paid Cogent to =
provide
>>> to you. Cogent's demand is unethical. They intentionally fail to
>>> deliver on the basic service expectation you pay them for and refuse
>>> to do so unless a third party to your contract also pays them.
>>=20
>> That's one way of looking at it.
>>=20
>> However, which of your transits don't bill for bits exchanged with =
other
>> customers of theirs...and how are they or you accounting for that =
traffic?
>=20
> Hi Jon,
>=20
> As you know, there is a technology limitation in how routing works
> which says that for any given block of addresses you can, absent
> extraordinary measures, have a peering relationship or a transit
> relationship but not both. If both parties choose to have a transit
Not really.
If you have both, then there=E2=80=99s no easy way to guarantee that you =
get
paid for every piece of transit (though relatively simple localpref
tactics will actually make it likely that you also get paid for
many bits of peering).
> relationship, that excludes a peering relationship for the relevant
> blocks of addresses. And that's OK when _both sides_ choose it.
Your premise is flawed.
> In related news, no ethical conundrum demands defiance of the law of =
gravity.
True, but gravity is real. Your law of peering vs. transit above is
purely artificial and fails utterly if you don=E2=80=99t accept that an =
approximation
of which bits fall into which category is =E2=80=9Cclose enough=E2=80=9D =
for billing
purposes.
I=E2=80=99m not making any value judgments on whether accepting that =
idea is good
or bad. I know that there are networks that act in various ways on both
sides of this idea.
However, equating it to =E2=80=9Cthe law of gravity=E2=80=9D is rather =
silly given that it
is 100% mutable if we take the accounting out of the picture.
No amount of monetary policy change can counteract gravity.
Owen
