To some extent that's true. However, as a counter-point, consider such
sites as sunsite, wustl, smc.vnet.net, etc. I doubt those sites would
continue to exist in a solely bandwidth sensitive pay-as-you-go world.
I think they count on flat rate connectivity to be able to continue
to exist. I don't think the elimination of those sites (and many others
like them) would benefit the net. Do you?
Owen
I'm not certain that they represent a true public service, as opposed
to simply interesting content. Interesting content can probably pay its
own way, even at retail prices. For example, the incremental cost to
send 10MB of data is only about 50 cents using normal retail rates [1].
Are you saying that whatever you're downloading isn't worth paying that?
(or watching the appropriate number of web ads, as I currently do to
download palm pilot apps, pc freeware, and today's weather gif?)
/John
[1] Retail T1 transit from major backbone, fully utilized, $2400 monthly,
presuming cost recovery over 4 peak hours of 20 business days = $30
per peak hour; each hour good for about 600 MB -> 5 cents / MB )
Are you saying that someone should be forced to pay for the privilege of
offering something for free to your customers? Things that your customers,
who I number among are requesting?
/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
Patrick Greenwell (800) 299-1288 v
Systems Administrator (925) 377-1212 v
NameSecure (925) 377-1414 f
\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/
There's no way to know what is being offered "for free", "in
exchange for ads", "due to a credit card transaction", etc.
It's all just traffic.
Customers who receive traffic currently bear some of the costs
and the sending customer bears some of the costs. In the case
of an off-net sender with shortest-exit routing and no offsetting
traffic in the other direction, the receiving customer ends up
bearing all of the costs.
/John
Are you saying that someone should be forced to pay for the privilege of
offering something for free to your customers? Things that your customers,
who I number among are requesting?
I don't know what he was saying, but I'll say something like it: gatekeeper,
wuarchive, cdrom.com, and other archives of free stuff are going to have to
do some kind of micropayment scheme -- charge for downloads in other words --
in order to pay their own costs to their providers, in order that those
providers are able to pay THEIR costs, in transit they buy or in glass they
lease or whatever.
The Internet backbone's growth has been all about barriers to entry and in
special deals. People have been buying their provisioning at flat rate or
with other subsidies, and reselling it at variable rates to folks who came
later or otherwise didn't have access to, or knowledge of, the special deals
of the pioneers. Eventually these special deals run out of time, or run out
of bandwidth, and a true (cost-driven) market economy is developing.
What we're seeing now is just SO inevitable.
>Are you saying that someone should be forced to pay for the privilege of
>offering something for free to your customers? Things that your customers,
>who I number among are requesting?There's no way to know what is being offered "for free", "in
exchange for ads", "due to a credit card transaction", etc.
Actually, since we were discussing sites such as sunsite, I think that we
at least have an idea. If we go to a settlement-based model, one can only
wonder what will happen to freeware/shareware sites.
It's all just traffic.
Sure it is. Traffic that the *receiver* is requesting.
Customers who receive traffic currently bear some of the costs
and the sending customer bears some of the costs. In the case
of an off-net sender with shortest-exit routing and no offsetting
traffic in the other direction, the receiving customer ends up
bearing all of the costs.
Well, my understanding is (and someone correct me if I am wrong) in at
least the case of Exodus, they aren't using closest-exit. I can completely
understand requiring peers not use closest-exit. That seems somewhat
reasonable. But it is still the receiver that is initiating the
transfer in these cases and "causing" the asymmetry. And the receiver,
someone like me, is paying for a bidirectional link to BBN already. I'm
already paying to receive the data, yet what you are proposing to do has
the sender paying as well.
I haven't seen anything in these recent discussions to suggest that BBN
would be offering me a discount on inbound traffic since now the sender
would be paying for it.
/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
Patrick Greenwell (800) 299-1288 v
Systems Administrator (925) 377-1212 v
NameSecure (925) 377-1414 f
\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/
See comments below....
Paul Vixie wrote:
> I don't know what he was saying, but I'll say something like it: gatekeeper,
wuarchive, cdrom.com, and other archives of free stuff are going to have to
do some kind of micropayment scheme -- charge for downloads in other words --
in order to pay their own costs to their providers, in order that those
providers are able to pay THEIR costs, in transit they buy or in glass they
lease or whatever.
I hate to ask, but are *you* buying into the host transmit charges?
As opposed to client request fee's?
I am *not* primarily a hosting site, but I don't *even* like that
concept. :\
(Not that anyone really cares about my *opinion*,
I just felt a need 
CDROM.COM could run PASV, and the "requester" would pay for BW.
(DS0m/s 
Hardly an unfair request. Now, if CDROM.COM has enough sponsers,
they could allow traditional FTP... (Changes the origin, and hence
fee's)
I could even imagine the JAVA app, that sets up the download,
popping up a little box that updates a small AD,( oh.. say every 30
seconds...) while you download via FTP from the other window...
Of course, that might hurt standard FTP programs, (or *help* them
"This download brought to you by....."
"Click here for a great deal on .... "
Just letting my imagination flow, pardon any obvious shortcomings..
> Are you saying that someone should be forced to pay for the privilege of
> offering something for free to your customers? Things that your customers,
> who I number among are requesting?I don't know what he was saying, but I'll say something like it: gatekeeper,
wuarchive, cdrom.com, and other archives of free stuff are going to have to
do some kind of micropayment scheme -- charge for downloads in other words --
in order to pay their own costs to their providers, in order that those
providers are able to pay THEIR costs, in transit they buy or in glass they
lease or whatever.
Ok, it hasn't happened yet, and software has been distributed free for
quite some time, but let's assume that this is indeed the inevitable.
Who pays?
Model 1 (current model):
a) Recipient pays provider extra for transmission.
b) Sender may/may not pay extra for transmission (transit vs. peering)
Model 2 (new, shiny, "improved" model):
a) Recipient pays sender for download.
b) Recipient pays extra for transmission.
c) Sender pays provider extra for transmission.
Well, wait a minute, doesn't the recipient already normally pay? I have
three different providers, BBN being one of them. All bill on the 95th
percentile. If I want more data, I pay for the privilege. Works for me. I
understand that some providers bill flat-rate. That's their business
decision, and may or may not scale.
In the case of BBN, I am paying for end to end connectivity to whatever
portion of the Internet that happens to be operational at that moment. *I*
am already paying for the cost of BBN to get the data from wherever to me.
So now under model 2, I am paying twice, not because cdrom.com or wherever
wants to charge me extra, but because they are forced to pay for the
privilege of sending me data that I requested and am already paying the
transmission costs for. The net effect is that the transit provider is
double-dipping, and it costs the receiver twice to receive the data.
Somehow this doesn't seem to be very attractive to me.
The Internet backbone's growth has been all about barriers to entry and in
special deals. People have been buying their provisioning at flat rate or
with other subsidies, and reselling it at variable rates to folks who came
later or otherwise didn't have access to, or knowledge of, the special deals
of the pioneers. Eventually these special deals run out of time, or run out
of bandwidth, and a true (cost-driven) market economy is developing.What we're seeing now is just SO inevitable.
Don't worry, we'll be paying for the commercial release of BIND
when you guys start selling it...
I thought the whole idea with this Net thingy was to make bandwidth so
cheap it wasn't an issue. This of course is a pipe dream. The real idea
from the "pros" seems to go something like: "The idea is to make
bandwidth so cheap that it doesn't cost us anything to deliver, but allows
us to charge the same or more for."
Am I close?
/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
Patrick Greenwell (800) 299-1288 v
Systems Administrator (925) 377-1212 v
NameSecure (925) 377-1414 f
\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/
John Curran wrote:
Customers who receive traffic currently bear some of the costs
and the sending customer bears some of the costs. In the case
of an off-net sender with shortest-exit routing and no offsetting
traffic in the other direction, the receiving customer ends up
bearing all of the costs.
I guess 'all the cost' means most of the cost, and 'no offsetting traffic'
means 'not much offsetting traffic'.
However, is the real problem here the traffic assymetry, or the fact that
all of the traffic is coming from one geographic location?
If it is the former, then there isn't much of a solution except to merge
with a network that sucks a huge amount of traffic. However, if it is the
latter, then wouldn't content distribution fix it? I know many web farms
offer distributed servers to their customers as a type of premium service.
However, since in this case it benefits all parties involved, it seems to me
that it might make sense to offer this service to huge web sites at little
or no additional cost.
Alec
Patrick Greenwell <patrick@namesecure.com> writes:
I don't know what he was saying, but I'll say something like it: gatekeeper,
wuarchive, cdrom.com, and other archives of free stuff are going to have to
do some kind of micropayment scheme -- charge for downloads in other words --
in order to pay their own costs to their providers, in order that those
providers are able to pay THEIR costs, in transit they buy or in glass they
lease or whatever.
Ok, it hasn't happened yet, and software has been distributed free for
quite some time, but let's assume that this is indeed the inevitable.
[...]
Model 2 (new, shiny, "improved" model):
a) Recipient pays sender for download.
b) Recipient pays extra for transmission.
c) Sender pays provider extra for transmission.
So now under model 2, I am paying twice, not because cdrom.com or wherever
wants to charge me extra, but because they are forced to pay for the
privilege of sending me data that I requested and am already paying the
transmission costs for. The net effect is that the transit provider is
double-dipping, and it costs the receiver twice to receive the data.
Somehow this doesn't seem to be very attractive to me.
But, it looks attractive to me. One of the problems with most of the
conceptual schemes posed is that they do not scale to the international
links, particularly across the Pacific, where the cost of bandwidth is
still very, very dear.
As a Korea based ISP, it is very much in my interest to bring content into
my network as close to the user as possible. When it enters in the US,
I pay the full cost (at ~U$16,000/Mbps for big pipes) to get it across
the lake.
There is little reason, in the current scheme, for me to transport
bozobanner.com's gif adverts across the pacific. My business is also
held hostage to any new wiz-bang push technology. And let's not even
get started on smurf attacks.
Given that my Korea domestic bandwidth costs are orders of magnitude
(about 1/160) times lower, my cost basis is extremely sensitive to the
amount of total international traffic. If the burden were placed on the
content provider, and through him on the requesting user, to pay for
that long haul, there would be a clear and present incentive for users
to find topologically closer content.
I sometimes toy with the idea of redirecting requests to some of the big
unsolicited content providers (default MS and Netscape pages, doubleclick,
etc.) to a local, unassociated site. Now, if they were somehow settling
for their traffic...or even if they created an supported a local mirror,
my costs of providing connectivity to them would be much reduced.
Also, with a large portion of my international traffic (and every on
else's, I presume) being porn, which uses so much bandwidth and has so
many auto-popup, unrequested adverts, this type of settlement starts
making more and more sense. Most of those sites are for pay, and some
of that income should make it back to the network doing the long haul.
The argument that my U$10/month dialup customers or U$1850/mo T1 leased
line customers should be entirely responsible for paying for this traffic
doesn't scale, and is indeed the evil of flat rate schemes taken to an
absurd extreme, with those who primarily access local content massively
subsidizing the porn freaks using international circuits.
This is the fundamental reason for the restricted bandwidth across big
ponds, and why no US backbone has any significant bandwidth of their
own across the Pacific.
The Internet backbone's growth has been all about barriers to entry and in
special deals. People have been buying their provisioning at flat rate or
with other subsidies, and reselling it at variable rates to folks who came
later or otherwise didn't have access to, or knowledge of, the special deals
of the pioneers. Eventually these special deals run out of time, or run out
of bandwidth, and a true (cost-driven) market economy is developing.
What we're seeing now is just SO inevitable.
Absolutely. Only the Australians have got anything close to proper, so far.
Don't worry, we'll be paying for the commercial release of BIND
when you guys start selling it...
I sure will.
I thought the whole idea with this Net thingy was to make bandwidth so
cheap it wasn't an issue. This of course is a pipe dream. The real idea
from the "pros" seems to go something like: "The idea is to make
bandwidth so cheap that it doesn't cost us anything to deliver, but allows
us to charge the same or more for."
It is a pipe dream, for the time being. It is not too difficult to see
a day when trans-US-continental DS0miles are cheap enough to make US
current speed dialup access (and by extention LD phone calls) nearly
free. But, of course, by then the bandwidth demands of most users will
expand to require something significantly more. And, we are no where
near that point in most international paths.
-jem
John Milburn jem@xpat.com
Director - BoraNet jem@bora.net
Cell +82 19-220-7035 Tel +82 2-220-7035
Dacom Corporation, Seoul, Korea Fax +82 2-220-0751
John Curran wrote:
Customers who receive traffic currently bear some of the costs
and the sending customer bears some of the costs. In the case
of an off-net sender with shortest-exit routing and no offsetting
traffic in the other direction, the receiving customer ends up
bearing all of the costs.I guess 'all the cost' means most of the cost, and 'no offsetting traffic'
means 'not much offsetting traffic'.
(yes)
However, is the real problem here the traffic assymetry, or the fact that
all of the traffic is coming from one geographic location?
Traffic Assymetry of traffic which requires (for lack of better term)
"transport" to the destination. In the case of longest-exit routing,
the traffic received doesn't require much transit, nor in the case of
distributed content.
If it is the former, then there isn't much of a solution except to merge
with a network that sucks a huge amount of traffic.
Presuming shortest-exit, correct.
However, if it is the latter, then wouldn't content distribution fix it?
I know many web farms
offer distributed servers to their customers as a type of premium service.
However, since in this case it benefits all parties involved, it seems to me
that it might make sense to offer this service to huge web sites at little
or no additional cost.
It's difficult to "know" whether the distribution is actually
working and resulting in interconnecti traffic which is local,
but otherwise, yes.
/John
...
Customers who receive traffic currently bear some of the costs
and the sending customer bears some of the costs. In the case
of an off-net sender with shortest-exit routing and no offsetting
traffic in the other direction, the receiving customer ends up
bearing all of the costs.Well, my understanding is (and someone correct me if I am wrong) in at
least the case of Exodus, they aren't using closest-exit. I can completely
understand requiring peers not use closest-exit. That seems somewhat
reasonable.
I was not referring to any particular peering relationship,
only problems brought about by closest-exit peering in the
presence of highly assymetric traffic.
I haven't seen anything in these recent discussions to suggest that BBN
would be offering me a discount on inbound traffic since now the sender
would be paying for it.
In the case of traffic coming coming from a peer network with wildly
asymmetric traffic, the sending network is paying to offset the traffic
assymetry; this returns the economics to that of a balanced peer or
an on-net sender (which is the normal case today).
/John
You're paying half the cost currently, the sender (presuming
they're transit or a relatively balanced peer) is paying the
other half.
/John
John Curran wrote:
It's difficult to "know" whether the distribution is actually
working and resulting in interconnecti traffic which is local,
but otherwise, yes.
Come on John, use your imagination. There are creative ways one can use
things like cisco flow-stats to figure out how much traffic is traversing
your cross-country links. As long as you too are doing closest exit routing
then it is possible to get some very good distribution of server load based
on the destination's location on your network. I spoke about one possible
solution at NANOG 11, http://www.hilander.com/nanog11. Of course, this
requires the web farms to actually deploy their huge clients using such a
content distribution scheme.
Alec
>>...
>> Customers who receive traffic currently bear some of the costs
>> and the sending customer bears some of the costs. In the case
>> of an off-net sender with shortest-exit routing and no offsetting
>> traffic in the other direction, the receiving customer ends up
>> bearing all of the costs.
>
>Well, my understanding is (and someone correct me if I am wrong) in at
>least the case of Exodus, they aren't using closest-exit. I can completely
>understand requiring peers not use closest-exit. That seems somewhat
>reasonable.I was not referring to any particular peering relationship,
only problems brought about by closest-exit peering in the
presence of highly assymetric traffic.
Well, again correct me if I am wrong, but I don't think you can discuss
specific relationships. That's understandable, but the actions of BBN in
this specific case aren't clear to me as it does not fit the profile you
are describing.
>I haven't seen anything in these recent discussions to suggest that BBN
>would be offering me a discount on inbound traffic since now the sender
>would be paying for it.In the case of traffic coming coming from a peer network with wildly
asymmetric traffic, the sending network is paying to offset the traffic
assymetry; this returns the economics to that of a balanced peer or
an on-net sender (which is the normal case today).
John, that does not at all address my statement. If the sender is paying
for requested data, is my bill going down? I'm fairly confident the
answer is no.
So this leaves a situation where the sender will most likely have to
charge me for the requested data as well in order to cover costs, and I
end up paying twice. I am not particularly fond of paying for things
twice. Once is quite enough thanks.
/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
Patrick Greenwell (800) 299-1288 v
Systems Administrator (925) 377-1212 v
NameSecure (925) 377-1414 f
\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/