RE: Peering, Large ISPs, and You

See my comments below...

From: David Barak [mailto:thegameiam@yahoo.com]
Sent: Wednesday, July 25, 2001 1:14 PM
To: nanog@merit.edu
Cc: dan@netrail.net; jeblinton@corp.earthlink.net;
peeringresistance@yahoo.com
Subject: Peering, Large ISPs, and You

I'll interleave my comments.

Quoth Daniel Golding:

>The large ISPs have finally started to work together,
>to potentially exclude
>smaller providers. That isn't good.

are you arguing that the situation of the past several
years does NOT exclude smaller providers? Most of the
really big ISPs (you know who you are) rely primarily
on private peering already. How exactly does this
change matters?

Smaller providers have certainly felt the squeeze. This changes matters
because previously the big players were working seperately, with distinct
sets of rules. From my reading of these comments, it looks like they are
starting to work together, and plan jointly. If this is true (and at least
one non-anonymous poster has seemed to confirm it), than thats bad.

>Certain colo facilities are being choosen. Others are
>not. This has a major
>business impact on the ones who aren't choosen.

s 'are being'/'had been' and you'll see that this is
not only nothing new, it's nothing which is
particularly undesireable. In fact, you can look at
this as colo-competition - those colo spaces which
provide services which the really big ISPs want will
get their business.

Yes. But previously, all these ISPs made independent decisions, which
encouraged competition. If all of them jointly choose a single player in a
certain market, thats bad for competition. That has NOT happened previously.

>Earthlink is a huge consumer of transit bandwidth, so
>it would seem to be in
>your shareholder's best interest to keep competition
>high, and thus keep
>prices low.

True. So tell me this: how will providers reducing
their costs on settlement-free interconnections cause
overall costs to rise? If anything, this period of
severe cost-sensitivity should drive the really big
ISPs to pay very close attention to pricing, in an
attempt to maintain and maximize revenue.

There are numerous examples of this in real life. The best one if the
airline industry. At any rate, once the pricing variability is taken out of
the transit game, syncing of pricing will happen. There seems to be a hope
that this will supress pricing. This may happen, although it is difficult to
see how transit can be delivered, at a profit, for lower prices. But be
careful to see the other side of the coin - it makes price increases much
easier.

-David Barak
"Quis custodes ipsos custodiet?" - Juvenal

__________________________________________________
Do You Yahoo!?
Make international calls for as low as $.04/minute with Yahoo! Messenger
http://phonecard.yahoo.com/

- Dan

sets of rules. From my reading of these comments, it looks like they are
starting to work together, and plan jointly. If this is true (and at least
one non-anonymous poster has seemed to confirm it), than thats bad.

Dan,

  this secret "cabal" of all controlling, all knowing large ISP's working
in concert to the detriment of the "small" ISP is bullshit. The processes
and politics are so bad that chances of a _single_ ISP's divisions making
a coherent plan are far more likely than developing a lighter than air
pig. Now apply that across 10 of these ISPs.

Yes. But previously, all these ISPs made independent decisions, which
encouraged competition. If all of them jointly choose a single player
in a certain market, thats bad for competition. That has NOT happened
previously.

This solution mandates many things, including entrance facilities, the
provider being willing to work with the fiber folks at pulling in the
strands, providing the power, space and cooling. Security, cross connect,
and remote hands. Given the state of the colo market, I am surprised
anyone got the business, and chances are, to get the business someone had
to be very willing to commit to meeting the requirements and so they got
the business. Capitalism at its finest as it were.

/vijay

Having done time in the US Military and had both employment and consulting
in and with large organization, some of them in this industry, I would
tend to be in wholesale agreement with Vijay. Large organization don't
conspire well, not even within themselves. Heck, it's hard to get more
than seven people to agree on anything. I worked at one company where the
board of directors had a two month argument about office stationary design.

I do think it would lance a few boils if the nine locations where released
to the community.