- NANOG - lists.nanog.org

Re: UUNET peering policy
by Sean Donelan 15 Jan '01

15 Jan '01

On Mon, 15 January 2001, Steve Meuse wrote: > Then you have not made the same investment in infrastructure, and therefore > are not a *peer*. That sounds like the same argument AT&T made when MCI tried to connect to its network. AT&T currently has a market capitalization of $91 Billion MCI/Worldcom currently has a market capitalization of $63 Billion Sprint currently has a market capitalization of $25 Billion Genuity currently has a market capitalization of $5 Billion Yep, I would say Genuity is not a peer of either AT&T, Sprint or Worldcom. Interconnection should be based on technical performance standards, not on the size of your balance sheet. I understand writing a performance based requirement is much harder than a prescriptive based requirement. Prescriptive: You will use OC-12 circuits. Performance: You will be able to carry 150 Mbps of sustained traffic for 5 minutes with less than 1% packet loss. If you want to use OC-12 circuits, great. If you want to use ATM, great. If you want to use quantum transfer, great. What I object to is the requirement to make the "same" investment, i.e. make the same mistakes. Perhaps I wanted to make my investment in circuits between the US and South America, so I have 150Mbps to Rio. This is essentially the argument that Telstra has been making to the UN and ITU. Telstra makes very large investments in its part of the world, some might argue even more than Worldcom invests. But because Telstra doesn't sell service in Virginia, none of that investment "counts."

1 0

Re: UUNET peering policy
by Sean Donelan 15 Jan '01

15 Jan '01

In the past, some providers have specified peering requirements such as "must use a Cisco 7000 with SSP". If you are concerned about congestion, you should specify a performance requirement covering congestion, not some internal architecture design of the other network. How the other network chooses to solve the congestion problem should be up to them. By saying you should treat the other provider as a black box I mean you should be able to tell if the other provider is meeting or not meeting their performance requirements with examination of how much they are paying for their circuits, what type of routers they are using, or how they design their internal network. On Sun, 14 January 2001, Paul Vixie wrote: > > ... I think interconnection agreements should be based on the point of > > interconnection. When you delve too much in how the internals of other > > providers' networks work, I think you are always going to run into > > problems. I think it is best to view other provider's networks as a > > black box. > > I don't agree. Peering is a business relationship transcending locations. > The best peering agreements I know of (e.g., mine) specify growth terms so > that traffic won't eventually encounter congestion while trying to get from > one to the other or back again. Some forms of growth involve additional > peering locations, and some do not. A peering agreement which was based on > the point of interconnection would be far less useful in avoiding congestion.

1 0

Re: UUNET peering policy
by Sean Donelan 15 Jan '01

15 Jan '01

On Sun, 14 January 2001, Paul Vixie wrote: > > ... The imbalance issue has come up a few more times with other > > providers such as PSI, Abovenet and others. > > To the best of my knowledge, AboveNet has never insisted on any particular > traffic balance with any of our peers. Send to us 10:1, 1:10, 1:1, whatever. > Any traffic coming or going over a peering connection is to or from one of > AboveNet's customers, which means we're paid (by that customer) to deliver it. > (Any other policy amounts to wanting to be paid twice for the same packet.) There are always two sides to any peering exchange. Abovenet may not insist on any particular traffic balance, but their peers may. So peering requirements may affect a provider because it is on the insister side or the insistee side. Abovenet is a nice example of this, simply because they are one of the few providers still publishing traffic statistics. (This is a good thing). If you look at Abovenet's traffic graphs, you'll notice Abovenet has a wide variety of traffic balances with different providers. Some in Abovenet's favor (such as 3:1 with Sprint, 5:1 with Teleglobe) and some in the other provider's favor (such as 1:3 with Exodus). And a whole bunch of providers somewhere in between. UUNET isn't listed, so I don't know what that traffic balance looks like. It will be interesting to see what happens in 12 months when UUNET retroactively applies their policy to existing private interconnections. What if you are a web hosting company with data centers in a few large cities (chi, dal, la, nyc, sf) and don't meet UUNET's requirement to be located in 15 US states. What if you are a major Canadian provider with POPs in every province from coast to coast, but only a few locations across the border in the USA. What if you are a major South American or African provider covering those entire continents, but with little presence in UUNET's strongholds of US, Europe and Asia. One problem with all these peering policies is different providers have different strengths. I think interconnection agreements should be based on the point of interconnection. When you delve too much in how the internals of other providers' networks work, I think you are always going to run into problems. I think it is best to view other provider's networks as a black box.

2 1

RE: UUNET peering policy
by Gregory Soo 15 Jan '01

15 Jan '01

Mediaspeak: Worldcom's Exclusive Peering Club Posts Rules http://www.zdnet.com/intweek/stories/news/0,4164,2674325,00.html By Max Smetannikov, Interactive Week, January 14 2001 8:19 PM ET ...The result will likely be a wave of consolidations among smaller access providers and a codified first and second class of data carriage, industry experts said... Interview: Vinton Cerf On Private Peering http://www.zdnet.com/intweek/stories/news/0,4164,2673857,00.html By Max Smettanikov, Interactive Week, January 15 2001 7:43 AM ET ...So I hope that when people look at this, they will understand that, especially in the case of private peering, when two parties have to build the physical capacity to link together, as opposed to public peering, when you have capacity going to a public NAP or MAE or what have you, it is even more clear evidence of cost associated with it. So it helps people understand the real framework in which Internet connectivity works... -----Original Message----- From: smd(a)clock.org [mailto:smd@clock.org] Sent: Monday, January 15, 2001 9:17 AM To: nanog(a)merit.edu; vixie(a)mfnx.net Subject: Re: UUNET peering policy Vixie writes: | Peering is a business relationship transcending locations. This is beautifully put, and I couldn't agree more. Each party sells the other connectivity of essentially unlimited bandwidth over time, but limited by scope. The prices for this service and associated committments to one another are agreed to be so close that actually exchanging money is unworthwhile. Sean.

1 0

Hotel Room Block Rate for NANOG 21 Expires Jan. 17
by Carol Wadsworth 15 Jan '01

15 Jan '01

*Reminder* If you are planning to attend the February NANOG meeting, the hotel room block discounted rate expires this Wednesday, January 17. URL: http://www.nanog.org/mtg-0102/hotel.html Sheraton Atlanta Hotel 165 Courtland at International Boulevard Atlanta, Georgia 30303 In-house Reservations: (404) 659-6500 Toll-Free In-house Reservations: (800) 833-8624 Room Block Name: NANOG Dates Available: February 17, 18, 19, and 20 Rate: $149/single; $149/double Cut-off date: January 17 Note: Please make your hotel reservation between the hours of 8:00 a.m. and 7:00 p.m. (EST). After 7:00 p.m. (EST) calls are forwarded to a central (nationwide) reservation line for Sheraton hotels and you may receive incorrect information regarding the NANOG room block.

1 0

Re: UUNET peering policy
by smd＠clock.org 15 Jan '01

15 Jan '01

Vixie writes: | Peering is a business relationship transcending locations. This is beautifully put, and I couldn't agree more. Each party sells the other connectivity of essentially unlimited bandwidth over time, but limited by scope. The prices for this service and associated committments to one another are agreed to be so close that actually exchanging money is unworthwhile. Sean.

1 0

Re: UUNET peering policy
by Paul Vixie 15 Jan '01

15 Jan '01

sean(a)donelan.com (Sean Donelan) writes: > If you look at Abovenet's traffic graphs, you'll notice Abovenet has a wide > variety of traffic balances with different providers. Some in Abovenet's > favor (such as 3:1 with Sprint, 5:1 with Teleglobe) and some in the other > provider's favor (such as 1:3 with Exodus). ... "Favor"? What, precisely, connotes "favor" in this regard? Sending more, or receiving more? And: why? > ... I think interconnection agreements should be based on the point of > interconnection. When you delve too much in how the internals of other > providers' networks work, I think you are always going to run into > problems. I think it is best to view other provider's networks as a > black box. I don't agree. Peering is a business relationship transcending locations. The best peering agreements I know of (e.g., mine) specify growth terms so that traffic won't eventually encounter congestion while trying to get from one to the other or back again. Some forms of growth involve additional peering locations, and some do not. A peering agreement which was based on the point of interconnection would be far less useful in avoiding congestion. -- Paul Vixie <Paul.Vixie(a)MMFN.COM> CTO and SVP, MFN (NASDAQ: MFNX) AboveNet, PAIX, and MIBH are subsidiaries of Metromedia Fiber Network, Inc.

1 0

I wonder if this is why above.net took the action it did
by Henry R. Linneweh 15 Jan '01

15 Jan '01

http://www.theregister.co.uk/content/6/16062.html Spammer wrecks UUNet email service -- Thank you; |--------------------------------| | Thinking is a learned process. | | ICANN member @large | | Gigabit over IP, ieee 802.17 | | working group | | Resilient Packet Transport | |--------------------------------| Henry R. Linneweh

1 0

Re: UUNET peering policy
by Paul Vixie 14 Jan '01

14 Jan '01

sean(a)donelan.com (Sean Donelan) writes: ...another stunningly well researched, accurate and articulate article, including the following: > ... The imbalance issue has come up a few more times with other > providers such as PSI, Abovenet and others. To the best of my knowledge, AboveNet has never insisted on any particular traffic balance with any of our peers. Send to us 10:1, 1:10, 1:1, whatever. Any traffic coming or going over a peering connection is to or from one of AboveNet's customers, which means we're paid (by that customer) to deliver it. (Any other policy amounts to wanting to be paid twice for the same packet.) Of course I can't commend on PSI or any of the other companies Sean mentioned. -- Paul Vixie <Paul.Vixie(a)MMFN.COM> CTO and SVP, MFN (NASDAQ: MFNX) AboveNet, PAIX, and MIBH are subsidiaries of Metromedia Fiber Network, Inc.

2 1

RE: How does one make not playing nice with each other scale? (Wa s: net.terrorism)
by Mathew Butler 14 Jan '01

14 Jan '01

I'll point out that someone banning a site for AUP violations should only be able to affect their own network, since policies are supposed to be organization-level. (Thus, inflicting your policy decision on someone else is a very gray area, legally and ethically.) The reason why MAPS-RBL works (and is legally protected) is because everyone who uses it must -consent- to using it, and take positive action on their end to configure it. This means that each organization generates and enforces their own policy, though with assistance from an outside consultancy. However, in this case, AboveNet is inflicting policy decisions on transit routes =without consent=. (It's obvious this is without consent, otherwise this thread would never have come up.) It brings up an interesting point, but the law has held in the past that boycotts are only effective and legal if they're voluntary, and if coercion is involved then it become "intimidation tactics" of "organized crime". I can't for the life of me imagine how the Internet is any different -- maybe if you tried to apply turnpike or private highway rules to it, but those rules are generally based on consent, as well. Just some thoughts... -Mat Butler -----Original Message----- From: Mark Mentovai [mailto:mark-list@mentovai.com] Sent: Saturday, January 13, 2001 10:26 AM To: Paul Vixie Cc: nanog(a)merit.edu Subject: Re: How does one make not playing nice with each other scale? (Was: net.terrorism) There are a few things that would stand in the way of adoption of something like this: first, each anti-route would require manual configuration, and that comes with its own set of problems. Another potential issue (this is purely theoretical, I'm not referring to any past, present, or future situation in particular) is that providers trying to blackhole a certain site for AUP violations may want to negatively impact reachability as much as possible, rather than purely keeping the offending traffic off their network. These folks wouldn't want to advertise anti-routes because the resulting blackhole avoidance would encourage others to take working alternate paths, which does less harm to the site in question. Still, this may be a beneficial, even if little-used, addition. Thoughts? Mark

2 1