fight club :) richard bennett vs various nanogers, on paid peering

Richard Bennett richard at bennett.com
Wed Nov 25 08:54:50 UTC 2009


   Thank you for your insights.
   Richard A Steenbergen wrote:

On Tue, Nov 24, 2009 at 10:00:52PM -0800, Richard Bennett wrote:


I haven't found a good source who knows what's going on outside his own
network.


Mr. Bennett,

You know when I first read your post, I assumed you were just ignorant
and confused about the topic of peering on the Internet. Then I saw you
actively refusing to listen to intelligent feedback by some of the most
experienced network operators and peering managers in the industry,
dismiss any idea that you didn't agree with as part of the "Google
conspiracy", and further embarrass yourself with comments which proved
you lacked understanding of even the most basic concepts of peering or
inter-network traffic exchange. Normally I would just write you off as
another Dean Anderson style nutjob, but I'm afraid that your ramblings
are so wrong and your closed-mindedness is so severe that you are
actually dangerous to anyone who might happen to read your comments and
think that they are in any way correct. Therefore, I think it is
important for all of us that you be refuted.

I'll start with a few points from your post and comments. You said:



I'm not sure that your 'on-net routes' is the same product as the Paid
Peering that Norton is interpreting; the Arbor study found a large
increase in the traffic that moves through these transit bypass paths,
and that's the actual story. While this service may have been
available for a while, its use is radically increasing. That's data,
BTW, not anecdote, so if you have a problem with the Arbor data,
you'll need some data of your own to refute it.


For starters, if you aren't sure what "on-net routes" and "paid peering"
even are, maybe you shouldn't be trying to comment on them. Second, the
Arbor study said absolutely NOTHING about an increase in traffic that
moves via peering vs transit, to say nothing of paid vs settlement free
peering. Arbor is completely and totally unable to identify anything
about money exchanged for bits in general, and from a technical
perspective there is absolute no difference between a paid and non-paid
peering.

You seem to be convoluting the purported "increase in traffic between
tier 2 networks" with a completely absurd belief that all traffic
between tier 1's was transit and all traffic between tier 2's is
peering. In reality, tier 2's routinely buy from and sell to each other,
peer with some tier 1's, and sell paid peering between themselves when
the business opportunities arise.

You later go on to state:



The Arbor study is evidence that traffic is shifting, and the
carrier-neutral peering site managers I've spoken with tell me they're
making something like 300 cross-connects a month. Do you think all
those cross-connnects are implementing settlement-free peering or
conventional transit agreements? I'm surmising that they aren't.


You have absolutely no basis to make the determination about what
percentage of the crossconnects are peering and what percentage are
transit. This is what we tried to explain to you with the "you can't
know this about any network but your own" answer, which you seemed
completely incapable of understanding. The reality is that no one can
know the answer for anything but themselves. For my network, I'd say
much less than 20% of our crossconnects are peering, with the vast
majority being customers, and a significant amount being intra-network
capacity (intra-pop, metro, and long-haul circuits) and transit. The
number may vary between networks, but again you have absolutely zero
basis to make any kind of claim about peering let alone settlement-free
vs paid based on the number of crossconnects in a colo.

Most of the other arguments are either meaningless or fall apart once
you remove some of the fundamental misunderstandings above, but there
are still plenty of other things which are completely absurd. For
example, you said:



Paid peering is a better level of access to an ISP's customers for a
fee, but the fee is less than the price of generic access to the ISP
via a transit network. The practice of paid peering also reduces the
load on the Internet core, so what's not to like? Paid peering
agreements should be offered for sale on a non-discriminatory basis,
but they certainly shouldn't be banned.


Paid peering (or peering of any kind) is absolutely no guarantee of
"better" access to any network, nor is it guaranteed (or even likely) to
reduce costs. There is also no such thing as "load on the Internet core"
to reduce, and this further illustrates a complete failure to understand
how the Internet works in general.

Paid peering is simply another form of transit, where two networks agree
to exchange money for the service of delivering connectivity. The only
difference is that you're only selling a portion of the routing table
rather than the "whole thing", for a specific subset of routes which
have different properties than the rest. In the case of paid peering,
the different property is that you'll get to bill your customer on the
other side for the traffic, thus allowing you to "double dip" for the
same bit and potentially make more money.

Of course in practice it doesn't work this way at all. The vast majority
of the cost of operating a network is transporting the bits from one
place to another, and when you sell paid peering you are guaranteed that
the traffic is going to stay on your network and be hauled. This makes
it some of the most expensive traffic to deliver, and typically results
in prices which are higher than those of another network who is hot
potatoing those bits off their network in one location, and who is
sending the traffic to a settlement-free peer. There is nothing wrong
with paid peering, it often has a time and a place (such as when two
networks are close to being settlement-free peers, but not quite, and
someone needs to sweeten the deal a little bit), but it is not the
panacea you think it is. Of course nobody else seems to think the FCC
Question 106 is talking about regulating paid peering (which would be
absurd), so fortunately I don't think we have anything to worry about.

Of course all of these points (and more) were already quite elegantly
expressed by fine folks like Vijay Gill, Dan Golding, Patrick Gilmore,
Joe Provo, and others. They tried to help correct your misinformation
with free advice, and you repaid them with delusional rants. Now you
simply look like a fool to everyone.



--
Richard Bennett
Research Fellow
Information Technology and Innovation Foundation
Washington, DC



More information about the NANOG mailing list