Re: Provider credibility - does it matter? was Re: Inter-provider relations
Karl Denninger <karl@Mcs.Net> wrote:
Any provider that does not recognize the value of bilateral, no-settlement peering anywhere that its cost-effective for both parties (ie: if you have traffic destined for me, get it on MY network where I'm being paid to carry it and let ME figure the rest out!) deserves what they get.
Zero-settlement peerings open to anyone are demonstrably amount to subsidies from large peers to small. That already was beaten to death. However, i repeat the argument: Big Provider Customer A ---[POP] ------------- 1000 miles -----------[POP] | IXP | Customer B ------[POP]-1 mile-[POP] Small Provider When customers A and B talk Big Provider pays to get them through 1000 miles. Small Provider pays for 1 mile. Note that i didn't even talk about less measurabe, but way too more important things like hosting of information suppliers. Say, Big Provider connects 1000 web sites; Small Provider hosts 1 site -- benefit from peering in terms of Web site diversity to the Big Provider's customers is 0.1%. To Small Provider's customers the benefit of peering is 99.9%. Zero-settlements work only when peers are of comparable size. Any attempt to extort pressure to force it upon anyone simply causes large folks to flee. --vadim
We have the same picture there. If we have 1000 kilimoters back-bone, we include the cost of this back-bone into our prices, and sell 64K for (for example) 100bokazoids. If small ISP crinix opens free-of-chsarge peering with us, he must not include the cost of back-bone into their prices, and they sell 64K for 50 bokazoids. This mean we can't allow free-of-charge peering with them.
Karl Denninger <karl@Mcs.Net> wrote:
Any provider that does not recognize the value of bilateral, no-settlement peering anywhere that its cost-effective for both parties (ie: if you have traffic destined for me, get it on MY network where I'm being paid to carry it and let ME figure the rest out!) deserves what they get.
Zero-settlement peerings open to anyone are demonstrably amount to subsidies from large peers to small.
That already was beaten to death. However, i repeat the argument:
Big Provider Customer A ---[POP] ------------- 1000 miles -----------[POP] | IXP | Customer B ------[POP]-1 mile-[POP] Small Provider
When customers A and B talk Big Provider pays to get them through 1000 miles. Small Provider pays for 1 mile.
Note that i didn't even talk about less measurabe, but way too more important things like hosting of information suppliers. Say, Big Provider connects 1000 web sites; Small Provider hosts 1 site -- benefit from peering in terms of Web site diversity to the Big Provider's customers is 0.1%. To Small Provider's customers the benefit of peering is 99.9%.
Zero-settlements work only when peers are of comparable size. Any attempt to extort pressure to force it upon anyone simply causes large folks to flee.
--vadim
--- Aleksei Roudnev, Network Operations Center, Relcom, Moscow (+7 095) 194-19-95 (Network Operations Center Hot Line),(+7 095) 239-10-10, N 13729 (pager) (+7 095) 196-72-12 (Support), (+7 095) 194-33-28 (Fax)
We have the same picture there.
If we have 1000 kilimoters back-bone, we include the cost of this back-bone into our prices, and sell 64K for (for example) 100bokazoids.
If small ISP crinix opens free-of-chsarge peering with us, he must not include the cost of back-bone into their prices, and they sell 64K for 50 bokazoids.
This mean we can't allow free-of-charge peering with them.
On the contrary. You can reach many cities, and manysets of customers with that 1000 km of backbone. The other guy cannot do so, and thus is inherently limited in his abilities to serve those same markets. You therefore inherently have more opportunity in the market (and more cost). If you did not think this was true, why did you build that 1000 kilometer backbone? Nobody FORCED you to do so, and to try to say "well, now the REST of the Internet gets to pay for MY business decisions" is extortion, pure and simple. You wish to twist around the logic to say that you "can't" peer for free with the other provider involved. Baloney. Let's look at this from another angle -- the one that matters. I'm your customer. I want to talk to one of that other provider's customers across town, and both of you are at MAE-East-Buttafuco. You want to charge the other provider to peer based on traffic (settlements). He does not want to charge you, because he thinks you both get equal value (you both have a customer involved who IS paying each of you, respectively) from the bilateral traffic. He refuses to pay or establish a transit connection. The EFFECT is that you screw ME. I'm paying YOU. The second-order effect is that you LOSE me as a customer, because when I can't reach that point and call your NOC, what are you going to tell me? That the other provider won't pay you? My answer is "*I* am paying you, and *I* want to talk to that network. Either provide the access or I will find a supplier that will AND STOP PAYING YOU." Now you have fewer customers to pay for that 1000 km backbone. Soon, if you keep it up, you won't have ENOUGH customers to pay for that 1000 km backbone. In fact, each customer who you screw and walks raises the pressure on you to try to get others to pay you to peer -- and increases the likelihood that you'll be told to stuff it time and time again. You're in a feedback loop which will lead to the destruction of your business. We're supposed to be running the INTERNET here, not the Balkan-net. -- -- Karl Denninger (karl@MCS.Net)| MCSNet - The Finest Internet Connectivity http://www.mcs.net/~karl | T1's from $600 monthly to FULL DS-3 Service | 23 Analog Prefixes, 13 ISDN, Web servers $75/mo Voice: [+1 312 803-MCS1 x219]| Email to "info@mcs.net" WWW: http://www.mcs.net/ Fax: [+1 312 248-9865] | 2 FULL DS-3 Internet links; 400Mbps B/W Internal
Karl the xenophove, gung-ho, anti-communist wrote:
We're supposed to be running the INTERNET here, not the Balkan-net.
'nuff said I think. -- Peter Galbavy peter@wonderland.org @ Home phone://44/973/499465 in Wonderland http://www.wonderland.org/~peter/
Its like two animals meeting in the jungle at night: If you are of equal "size" you'll quite happily sit about, exchange pleasantries, and even peer IP packets with zero dollar settlement. But if either creature is bigger than the other then you can expect a somewhat different outcome, normally involving a forced exchange of protein. But the problem is that you have to work out who is bigger, or work out if you are both of the same size, and all you can see is these two little yellow dots for eyes right in front of you in the gloom.... Zero dollar peering is not a panacea for Internet connectivity: If party A invests a few squillion dollars in infrastructure it is unlikely that they will peer with party B who has purchased a PC and a couple of modems. Normally Party A would say to B - "tough luck sport, it s a client relationship we are talking about here", simply as the issue here is that if A offers 0 dollar peering to B, A writes off a small part of their investment. This is not an economically stable relationbship. Stability comes when A and B percieve that the exchange of traffic is of equal benefit to both parties, and in general this happens when A and B are of equal "size". Thanks, Geoff
We have the same picture there.
If we have 1000 kilimoters back-bone, we include the cost of this back-bone into our prices, and sell 64K for (for example) 100bokazoids.
If small ISP crinix opens free-of-chsarge peering with us, he must not include the cost of back-bone into their prices, and they sell 64K for 50 bokazoids.
This mean we can't allow free-of-charge peering with them.
Karl Denninger <karl@Mcs.Net> wrote:
Any provider that does not recognize the value of bilateral, no-settlement peering anywhere that its cost-effective for both parties (ie: if you have traffic destined for me, get it on MY network where I'm being paid to carry it and let ME figure the rest out!) deserves what they get.
Zero-settlement peerings open to anyone are demonstrably amount to subsidies from large peers to small.
That already was beaten to death. However, i repeat the argument:
Big Provider Customer A ---[POP] ------------- 1000 miles -----------[POP] | IXP | Customer B ------[POP]-1 mile-[POP] Small Provider
When customers A and B talk Big Provider pays to get them through 1000 miles. Small Provider pays for 1 mile.
Note that i didn't even talk about less measurabe, but way too more important things like hosting of information suppliers. Say, Big Provider connects 1000 web sites; Small Provider hosts 1 site -- benefit from peering in terms of Web site diversity to the Big Provider's customers is 0.1%. To Small Provider's customers the benefit of peering is 99.9%.
Zero-settlements work only when peers are of comparable size. Any attempt to extort pressure to force it upon anyone simply causes large folks to flee.
--vadim
--- Aleksei Roudnev, Network Operations Center, Relcom, Moscow (+7 095) 194-19-95 (Network Operations Center Hot Line),(+7 095) 239-10-10, N 13729 (pager) (+7 095) 196-72-12 (Support), (+7 095) 194-33-28 (Fax)
Why did not you use the example "Bird and Alligator"? They have different sizes, but their meeting do not cause _the protein exchange_. -:) Seems there is 2 different cases (in ISP example): (1) big ISP and small ISP working in the same market and on the same territory; (2) Big ISP in USA and other ISP in ZIMBAMVE. If Alligator meet the hear, it try to eat this _piece of protein_. If he'll meet a small bird - we would see the example of mutual cooperation. Why not? (may be I have to name it Crocodile? I am not shure). Really there is 2 theard in this discussion - (1) and (2).
Its like two animals meeting in the jungle at night:
If you are of equal "size" you'll quite happily sit about, exchange pleasantries, and even peer IP packets with zero dollar settlement.
But if either creature is bigger than the other then you can expect a somewhat different outcome, normally involving a forced exchange of protein.
--- Aleksei Roudnev, Network Operations Center, Relcom, Moscow (+7 095) 194-19-95 (Network Operations Center Hot Line),(+7 095) 239-10-10, N 13729 (pager) (+7 095) 196-72-12 (Support), (+7 095) 194-33-28 (Fax)
On Sat, 26 Oct 1996 alex@relcom.eu.net wrote:
If Alligator meet the hear, it try to eat this _piece of protein_. If he'll meet a small bird - we would see the example of mutual cooperation.
But if either creature is bigger than the other then you can expect a somewhat different outcome, normally involving a forced exchange of protein.
When you look at the relationship between two entire species like rabbits and wolves it begins to look even more like cooperation where the rabbits cooperate by providing food for the wolves and the wolves cooperate by NOT WIPING OUT THE ENTIRE RABBIT POPULATION! It's called balance. Michael Dillon - ISP & Internet Consulting Memra Software Inc. - Fax: +1-604-546-3049 http://www.memra.com - E-mail: michael@memra.com
Its like two animals meeting in the jungle at night:
If you are of equal "size" you'll quite happily sit about, exchange pleasantries, and even peer IP packets with zero dollar settlement.
But if either creature is bigger than the other then you can expect a somewhat different outcome, normally involving a forced exchange of protein.
Oh, not always by any stretch of the imagination. Trying to eat someone in this fashion, in the wild, is traditionally risky. You can get poisoned, you can fail, and you can get tricked into doing something that causes you to become the food. Just for openers.
Zero dollar peering is not a panacea for Internet connectivity: If party A invests a few squillion dollars in infrastructure it is unlikely that they will peer with party B who has purchased a PC and a couple of modems. Normally Party A would say to B - "tough luck sport, it s a client relationship we are talking about here", simply as the issue here is that if A offers 0 dollar peering to B, A writes off a small part of their investment. This is not an economically stable relationbship. Stability comes when A and B percieve that the exchange of traffic is of equal benefit to both parties, and in general this happens when A and B are of equal "size".
Thanks,
Geoff
No. The exchange of traffic is *always* of equal benefit to both parties. Both parties have a customer who wants to get to the other (if this was not the case there would be no flow of traffic at all). Each party has been PAID to get that traffic to the other. An attempt to get paid TWICE is extortionate and the act of a firm which is trying to throw its weight around. It MAY even be actionable. It *certainly* is good marketing fodder for the smaller firm in every case, and I will say this -- we can, do, and WILL exploit these irregularities. We talk to smaller providers all the time. Part of the game is making sure customers are INFORMED. Part of informing them includes making sure they know what the real choices are, who has what USABLE bandwidth, where it goes, and what *policy* decisions each provider might be making that would affect connectivity and quality of service. Deliberate refusals to interconnect are just another part of this. An INFORMED consumer will insure that the market balances itself out. The only way the "big providers" can play this kind of game is if the consumer is NOT informed. Ignorance is, in our experience, very easy to fix. -- -- Karl Denninger (karl@MCS.Net)| MCSNet - The Finest Internet Connectivity http://www.mcs.net/~karl | T1's from $600 monthly to FULL DS-3 Service | 23 Analog Prefixes, 13 ISDN, Web servers $75/mo Voice: [+1 312 803-MCS1 x219]| Email to "info@mcs.net" WWW: http://www.mcs.net/ Fax: [+1 312 248-9865] | 2 FULL DS-3 Internet links; 400Mbps B/W Internal
Karl Denninger <karl@Mcs.Net> wrote:
Any provider that does not recognize the value of bilateral, no-settlement peering anywhere that its cost-effective for both parties (ie: if you have traffic destined for me, get it on MY network where I'm being paid to carry it and let ME figure the rest out!) deserves what they get.
Zero-settlement peerings open to anyone are demonstrably amount to subsidies from large peers to small.
No they're not. The load which the small provider presents to you (in the form of traffic to your CUSTOMERS) is miniscule by comparison. Further, so is the load you present to THEM. Finally, and FAR more importantly, the REASON you're having the traffic dumped to you is that *YOUR CUSTOMER IS PAYING YOU TO GET IT TO THEM*. If you refuse to perform that job, then your customer should find someone who will actually live up to the letter and spirit of what your customer is purchasing from you.
That already was beaten to death. However, i repeat the argument:
Big Provider Customer A ---[POP] ------------- 1000 miles -----------[POP] | IXP | Customer B ------[POP]-1 mile-[POP] Small Provider
When customers A and B talk Big Provider pays to get them through 1000 miles. Small Provider pays for 1 mile.
So what? Customer A paid you to get the traffic to him. It is in your best interest to do it. You got paid to do this. If you can't, Customer A will find Big Provider #2 (or Small Provider #2) who will. The first time you tell a CUSTOMER as "Big provider" that "the reason you can't reach Customer B, who you think is important, is because they aren't connected to us and their provider won't *pay us to transport YOUR DATA* you are going to find out, quickly, what the Customer's response to that is. You might find out from their corporate counsel; if not, you'll definitely find out from their purchasing department (or person) -- when they cancel your service and move somewhere else. Second, if Small Provider who *does* have capacity to that exchange point finds out what you're up to, expect to have that widely used in press materials and marketing efforts.
Note that i didn't even talk about less measurabe, but way too more important things like hosting of information suppliers. Say, Big Provider connects 1000 web sites; Small Provider hosts 1 site -- benefit from peering in terms of Web site diversity to the Big Provider's customers is 0.1%. To Small Provider's customers the benefit of peering is 99.9%.
Not if you're a customer of Big Provider and want to get there. Your provider either PROVIDES or you find someone who will. You seem to forget the middle letter of ISP is *SERVICE*. You want to talk to someone with a valid IP address on the Internet, your PROVIDER is responsible for seeing that you can get to an exchange point where can be found the network that serves them. BOTH providers have an obligation here, and its not to each other. Its to their CUSTOMERS. If the receiving network then refuses to accept the traffic destined for a customer WHO IS PAYING THEM TO TRANSPORT IT, the customer on that end has a very legitimate beef with their provider and IMHO has every right to walk away and possibly even sue, contract or no (the provider just breached their material obligations).
Zero-settlements work only when peers are of comparable size. Any attempt to extort pressure to force it upon anyone simply causes large folks to flee.
--vadim
Anyone short-sighted enough who fails to understand that bilateral, no-settlement PEERING (*NOT* transit) is in everyone's best interest deserves what the market does to them. I presume that (1) the people you peer with are clueful and don't do stupid things on a regular basis, (2) they don't try to point default at you, etc. That's a given in these discussions. -- -- Karl Denninger (karl@MCS.Net)| MCSNet - The Finest Internet Connectivity http://www.mcs.net/~karl | T1's from $600 monthly to FULL DS-3 Service | 23 Analog Prefixes, 13 ISDN, Web servers $75/mo Voice: [+1 312 803-MCS1 x219]| Email to "info@mcs.net" WWW: http://www.mcs.net/ Fax: [+1 312 248-9865] | 2 FULL DS-3 Internet links; 400Mbps B/W Internal
At 8:36 PM -0700 10/24/96, Vadim Antonov wrote:
Note that i didn't even talk about less measurabe, but way too more important things like hosting of information suppliers. Say, Big Provider connects 1000 web sites; Small Provider hosts 1 site -- benefit from peering in terms of Web site diversity to the Big Provider's customers is 0.1%. To Small Provider's customers the benefit of peering is 99.9%.
Oops, this has an arithmetic fallacy. Assuming that Big Provider also has 1000x more customers than Small Provider, and that all 1001 web sites are equally attractive, then there are 1000 BP customers each with a .001 chance of wanting to surf the SP web site. Compare this with 1 SP customer with a .999 chance of wanting to surf a BP web site, and a .001 chance of surfing the SP web site. In both directions the average traffic is the same (1 surf per unit time); SP is getting the same information supply benefit as BP. Of course this will show up as equal traffic in both directions, so it would be a wash under settlements. To the extent that one side's sites are better for surfing (e.g. higher wavers :-) or has proportionally fewer customers, traffic will be unequal. Roger
At 8:36 PM -0700 10/24/96, Vadim Antonov wrote:
Note that i didn't even talk about less measurabe, but way too more important things like hosting of information suppliers. Say, Big Provider connects 1000 web sites; Small Provider hosts 1 site -- benefit from peering in terms of Web site diversity to the Big Provider's customers is 0.1%. To Small Provider's customers the benefit of peering is 99.9%.
Of course this will show up as equal traffic in both directions, so it would be a wash under settlements. To the extent that one side's sites are better for surfing (e.g. higher wavers :-) or has proportionally fewer customers, traffic will be unequal.
Roger
The volume of traffic flow in each direction is irrelavent. In all cases where we're talking about *peering* (not transit) the source or destination of the traffic under consideration is someone from whom you collect money to provide service to. You are *obligated*, by ethic if not by law, to provide the best service possible given your installed infrastructure and capability to do so. Asking a *third* party to pay you for something that you've *already been paid for* is unethical in the extreme and, to the extent that providers are trying this in the marketplace, they should be exposed to public scrutiny. These kinds of extortionist tactics need to be brought out in the light of day where everyone can see them -- especially the customers of the providers who are doing these things. Only through a full public disclosure and discussion of this practice will those provider's customers come to understand that their provider, on an ethical level at minimum, is *not* providing to them what they think they're paying for. -- -- Karl Denninger (karl@MCS.Net)| MCSNet - The Finest Internet Connectivity http://www.mcs.net/~karl | T1's from $600 monthly to FULL DS-3 Service | 23 Analog Prefixes, 13 ISDN, Web servers $75/mo Voice: [+1 312 803-MCS1 x219]| Email to "info@mcs.net" WWW: http://www.mcs.net/ Fax: [+1 312 248-9865] | 2 FULL DS-3 Internet links; 400Mbps B/W Internal
participants (8)
-
alex＠relcom.eu.net -
Edward Vielmetti -
Geoff Huston -
Karl Denninger -
Michael Dillon -
Peter Galbavy -
Roger Bohn -
Vadim Antonov