peering charges?

Mailman List Mirror (Read Only)
1

The big problem is the current incentive to cripple network connections.

The incentive to reduce quality down to the minimal level accepted by
the target customer base is the prominent feature of any market.

Why should it cost more to connect at 100MB/sec vs. 56KB/sec? If the same
amount of packets are transfered? Why should I pay more beyond the
one-time cost of buying interface electronics?

Because there is a real cost for long-haul packet transport.

Why people pay higher rent for houses than for studio apartments?

"Connection costs" is rent on transmission facilities, plus
overhead (upkeep of the property, insurance, administration, etc).

Many real-time problems with packet switching (out of order arrival of
packets, need to re-order packets) essentially go away when speed
increases, i.e. things move out of perceptible ranges into in-perceptible
time variations.

That does not parse. Internet as is does not reorder packets.

Increasing speed does not change anything in this respect. It only
shrinks time scale.

And, i never suspected that reordering is a problem in real-time world.
0.5 sec variance (a lot!) with 1.5Mbps MPEG-2 video stream (better-than
cable TV quality) requires stunning 100 kilobytes worth of buffers.

The only reason reordering is a problem is because TCP has no reasonable
selective ACKs (the existing SACK options are close to useless), so
kludges like VJ's fast retransmit are necessary on high delay-bandwidth
networks.

Settlement based on whoever takes more traffic would be nice.

Would you please describe any useful mechanism of traffic-based
inter-backbone settelemts?

It gives a very direct incentive to build better networks.

Nah. It is like saying that "usage-based" pricing at McDonalds
gives a very direct incentive to sell something better resembling food.

Lets say MAE's
and NAP's are run in a cost-recovery mode and my only monthly cost as a
network is the cost of my circuits, i.e. having a connection to the
MAE/NAP.

They do it now.

As soon as my network takes more traffic than my circuit costs
I'm making money.

I assume "my network takes more traffic" means "there are more packets
(bytes, etc) going into my network", right?

So, you want to charge others for your customers browsing their customer's
servers?

Money that can be used directly to build a better
network which will take even more traffic, i.e. generate more money. In a
sense, this would provide best case capitalism with routers deciding from
whom to buy.

I love it. Do you really think you discovered a fount of money? Any settlement
game is zero-sum by definition. When ISP A pays ISP B some sum, A has
less resources to invest into infrastructure. On average, customers
do not win.

All energy would go into building better networks... neither
advertising, nor salespeople are necessary in this scenario!

Ah, i'd like to live in that world of yours.

--vadim

2

<this was later but might be more interesting>

Would you please describe any useful mechanism of traffic-based
inter-backbone settelemts?

Each packet either

1. terminates in my network
2. transits my network
3. originates from my network

Lets assume that I announce a price that I will charge other networks for
transitting packets through my network. Note, that I couldn't care less
about the speed at which they connect to my network.

Computers (third party?) on the edges of my network count the number of
packets that transit my network over a certain time period. Periodically I
issue invoices. Of course, other networks do the same. End users
(including co-located computers) simply appear as connected networks that
don't provide any transit, hence they pay for their connection.

>Why should it cost more to connect at 100MB/sec vs. 56KB/sec? If the same
>amount of packets are transfered? Why should I pay more beyond the
>one-time cost of buying interface electronics?

Because there is a real cost for long-haul packet transport.

Remember that I said "if the same amount of packets are transfered". The
speed of my connection does not necessarily affect the number of packets
that I'm transmitting.

"Connection costs" is rent on transmission facilities, plus
overhead (upkeep of the property, insurance, administration, etc).

No. That's "transmission cost", not "connection cost". The cost of me
connecting to your network (not counting setup and so on) is the sum of
the interface electronics on both ends and whatever link we need
inbetween.

Off on a tangent... creating bandwidth is somewhat like making computer
chips. Making the first production Pentium (or whatever) processor costs
billions. The second is a few cents. Here, laying the cable is expensive.
Once it is in the ground, the cost for transporting another packet is
almost zero.

An interesting question is what do to with potential bandwidth/cpu
cycles/free ram, i.e. what are the costs of not using available capacity?
One answer is that the cost is zero until your competitor starts being
more efficient then you are. See Soviet UNION vs USA.

Back to networks... it would be logical to assume that whenever people
deploy fiber they drop as much as they can afford at that moment. If
that's true then there must be enourmous amounts of unlit fiber. What's
the total capacity going accross the atlantic or pacific? How much of it
is being used? I'd bet that there are many terabits of fiber that are not
in use.

Yes, of course, the closer you get to what's technologically feasable,
(the closer you move to the border to the future) the more expensive
things become. I guess that's the main argument used to defend higher
costs for faster connections.

But we don't even need to go to that edge. Building gigabit, if not
terrabit routers is amazingly simple and can be done with off the shelf
technology. Hey, a few Linux boxes interconnected with a few 100MB/sec
ethernet switches in the right fashion would allow the creation of a super
NAP that should outperform gigaswitches easily.

Summary: I would not be surprised if the packet carrying capacity of the
Internet could be increased by two or three orders of magnitude with
surprisingly little investments. The real challenge is how to get people
to do that.

>All energy would go into building better networks... neither
>advertising, nor salespeople are necessary in this scenario!

Ah, i'd like to live in that world of yours.

Hehe... you can, think about it.

Dirk