[186566] in North American Network Operators' Group
Re: interconnection costs
daemon@ATHENA.MIT.EDU (Bill Norton)
Wed Dec 23 16:42:07 2015
X-Original-To: nanog@nanog.org
From: Bill Norton <bill.norton@gmail.com>
In-Reply-To: <CACGuEhHMgP7WDGqOVGpX9L+KuihUMHBDwTAR78jhnsLD3itx9w@mail.gmail.com>
Date: Tue, 22 Dec 2015 15:13:13 -0800
To: Reza Motamedi <motamedi@cs.uoregon.edu>
Cc: "nanog@nanog.org" <nanog@nanog.org>
Errors-To: nanog-bounces@nanog.org
Hi Reza -=20
When researching the costs of peering you should perhaps categorize into =
the most popular forms of peering.
Public (many-to-many) peering solutions vs. private (one-to-one)
=
--------------------------------------------------------------------------=
------------
There are of course many opinions on the merits of public peering vs. =
private peering, economically, technically, and strategically, but the =
unit cost per bit varies in both cases based on how much traffic is =
exchanged. The unit cost is the cost of peering divided by the amount of =
traffic peered, giving us a measure in $/Mbps. Network operators often =
compare this against the unit cost of transit, and make their decisions =
based primarily on cost. Generally I have seen content and cloud =
companies care more about the end-user experience and less about the =
cost of delivering the bits. To them peering directly provides better =
connectivity, so even if it costs more that Internet Transit it is often =
strategic to do so to improve the end-user experience. We now have data =
to back this improved end-user experience.
Remote Peering=20
----------------------
And then consider that one can remove the capital costs and reduce the =
opex for public and private peering through a technique called remote =
peering (aka =E2=80=98tethering=E2=80=99) into the well populated IXPs.=20=
Here we can remove the cost of the routers (amortized to thousands per =
month typically), opex for powered colocation (maybe thousands per =
month) and deployment costs. The transport cost remains. One write up I =
did in The Internet Peering Playbook showed that remote peering into an =
IXP can be had for about the cost of the transport alone. I also =
collected the religious arguments from the field highlighting the =
arguments for and against remote peering. This can be found in the book =
as well as in this blog:
=
http://drpeering.net/AskDrPeering/blog/articles/Ask_DrPeering/Entries/2012=
/9/18_The_Great_Remote_Peering_Debate.html =
<http://drpeering.net/AskDrPeering/blog/articles/Ask_DrPeering/Entries/201=
2/9/18_The_Great_Remote_Peering_Debate.html>
I hope this helps -
Bill=20
> On Dec 22, 2015, at 11:11 AM, Reza Motamedi <motamedi@cs.uoregon.edu> =
wrote:
>=20
> Thanks guys for the replies.
>=20
> I wanted to clarify two things in my questions. First by peering I did =
not
> necessarily mean "settlement free" interconnection. I meant any =
inter-AS
> connection. My understanding is that in addition to the cost of =
transit
> that should be paid to the transit provider, there also exists the =
cost of
> the xconnect that is charged by the colocation provider. Secondly, my
> question was more about the expenses, as opposed to the technical
> costs/benefits. I have browsed through the "Peering Playbook", but I =
think
> its more about providing a case "settlement free" peering.
>=20
> Best Regards
> Reza Motamedi (R.M)
> Graduate Research Fellow
> Oregon Network Research Group
> Computer and Information Science
> University of Oregon
>=20
> On Tue, Dec 22, 2015 at 9:33 AM, James Bensley <jwbensley@gmail.com> =
wrote:
>=20
>> On 22 December 2015 at 16:44, Reza Motamedi <motamedi@cs.uoregon.edu>
>> wrote:
>>> I think there is no single answer as different businesses may have
>>> different pricing models. I hope the discussion can help me =
understand
>> the
>>> whole ecosystem a little bit better.
>>=20
>>=20
>> Hi Reza,
>>=20
>> I have a list of example items that need to be costed in below, it is
>> by no means a definitive list though:
>>=20
>>=20
>> =
https://docs.google.com/document/d/1i2bPZDt75hAwcR4iKMqaNSGIeM-nJSWLZ6SLTT=
nuXNs/edit?pref=3D2&pli=3D1#
>>=20
>>=20
>> Cheers,
>> James.
>>=20
>>=20
