Sean - We sign and comply with mutual non-disclosure agreements that inhibit my ability to share that information with you. This is not a technical issue. I look forward to viewing your backhoe reports. - jsb -- Jeff Barrows Director, Internetwork Engineering UUNET, an MCI-Worldcom Company
Date: 19 Apr 2000 13:30:38 -0700 From: Sean Donelan <sean@donelan.com> To: nanog@merit.edu Subject: Re: Peering Table Question
On Wed, 19 April 2000, Jeff Barrows wrote:
One shouldn't necessarily believe any third-party web pages, documents, articles, or verbal statements about which networks any given network is peered with.
Though I have seen many articles, web pages, and other tables that detail which networks are 'peering,' I have never seen an accurate representation of this type of data from a third party.
Then I hope you would step forward and show us the best practice for making accurate information available about which networks any given network peers with starting with yours. I look forward to viewing the information about your network peering on your website or IRR or whatever method you decide to use.
To someone troubleshooting a routing or performance problem, it might look operational. A large file of baseline traceroutes could replace the benefits of disclosure, I suppose. Lawyers - 1 Tekkies - 0 my .02 -ls- Jeff Barrows <jsb@UU.NET> wrote:
Sean -
We sign and comply with mutual non-disclosure agreements that inhibit my ability to share that information with you.
This is not a technical issue.
I look forward to viewing your backhoe reports.
- jsb
-- Jeff Barrows Director, Internetwork Engineering UUNET, an MCI-Worldcom Company
Date: 19 Apr 2000 13:30:38 -0700 From: Sean Donelan <sean@donelan.com> To: nanog@merit.edu Subject: Re: Peering Table Question
On Wed, 19 April 2000, Jeff Barrows wrote:
One shouldn't necessarily believe any third-party web pages, documents, articles, or verbal statements about which networks any given network is peered with.
Though I have seen many articles, web pages, and other tables that detail which networks are 'peering,' I have never seen an accurate representation of this type of data from a third party.
Then I hope you would step forward and show us the best practice for making accurate information available about which networks any given network peers with starting with yours. I look forward to viewing the information about your network peering on your website or IRR or whatever method you decide to use.
A large file of baseline traceroutes could replace the benefits of disclosure, I suppose.
if it could, many of us would beg someone to do it RIGOROUSLY and publish. a few problems: o routing is not always symmetric, or i guess rarely is o you would have to do it from'*many* distributed locations to many distributed locations to see all the peering points through the potato fog o heck, due to bgp best path propagation, unless you are on a router with which P is a peer, you can not really tell if C is a customer or peer of P. o and it still would not tell you who is paying whom when it is peering randy
At 05:28 PM 4/19/00 -0700, Randy Bush wrote:
A large file of baseline traceroutes could replace the benefits of disclosure, I suppose.
if it could, many of us would beg someone to do it RIGOROUSLY and publish.
[...]
o and it still would not tell you who is paying whom when it is peering
What has that got to do with *operational* benefits. And why would it make any difference to a traceroute?
randy
TTFN, patrick -- I Am Not An Isp - www.ianai.net ISPF, The Forum for ISPs by ISPs, <http://www.ispf.com> "Think of it as evolution in action." - Niven & Pournelle (Enable? We dunt need no stinkin' enable!!)
Sean -
We sign and comply with mutual non-disclosure agreements that inhibit my ability to share that information with you.
This is not a technical issue.
If you are unable to provide this information, then don't ask stupid, goading questions in the first place. You don't need to wave the "third parties are idiots" stick around too much, unless you have a better way of proving your marketing-driven claims. UUNET (or whatever the name is this week in the land of Dogbert) may claim to be the world largest ISP, but they don't make many friends along the way. Those who want the money ... line up here. Those who want respect of their peers ... over there. Sorry, we don't have a queue for both. Peter
I read through this entire thread and it has shown to hit a raw nerve causing some comments to get away from the professionals that wrote them. I dont consider any network I peer with "evil" although they may have policies and practices I wholly disagree with. 1st the thread was correct to state that each network should have a written policy for implementing/initiating peering with another network. I do not believe the gov coming in and making a blanket policy is in the best interest of the net, however that is a danger if some make initiating peering impossible to achieve. The way different backbones implement peering IS a differentiator. 2nd I think this traffic balance issue is ridiculous and always have. If you have eyeballs they are paying for connectivity to the net, specifically content. In the rest of telecom those that initiate the request get billed so here eyeballs should be billed... and they are by their ISPs. If the network infrastructure is so expensive to maintain that the eyeball ISP losses money then they changed the incorrect rate. 1 word here caching. If the eyeball backbone wants to save network capacity and costs dropping some caches and give Doug H a call and add Cidera feed. If your network design is such that you cannot provide this caching service you better plan for it on your next build. The obligation of the big content backbones/websites should be to have multiple sites or inverse caching at the exchange cities to eliminate an undo burden on the eyeball backbones. Fair is fair. Both types of backbones, the eyeballs and the content backbones need to work together. In ever peering doc Ive seen there is a section on working together to enhance each others networks etc etc. That is rarely done in good spirit that it was intended. That's a shame and the responsibility falls for the most part squarely on us professionals on this list. With DSL coming along wont it be interesting to see how this will put pressure on the peering issue. David At 4:01 PM +0100 4/25/00, Peter Galbavy wrote:
Sean -
We sign and comply with mutual non-disclosure agreements that inhibit my ability to share that information with you.
This is not a technical issue.
If you are unable to provide this information, then don't ask stupid, goading questions in the first place. You don't need to wave the "third parties are idiots" stick around too much, unless you have a better way of proving your marketing-driven claims.
UUNET (or whatever the name is this week in the land of Dogbert) may claim to be the world largest ISP, but they don't make many friends along the way.
Those who want the money ... line up here. Those who want respect of their peers ... over there. Sorry, we don't have a queue for both.
Peter
-- Thank you, David Diaz Chief Technical Officer Netrail, Inc email: davediaz@netrail.net, davediaz@fla.net, cougar@mail.rockstar.org pager: davediaz@bellsouthips.com NOC: 404-522-1234 Fax: 404 522-2191 ----------------------------------- Build 1: 46 cities nationwide -- COMPLETE Build 2: 80 OC48s Nationwide [no typo] ++ FAILURE IS NOT AN OPTION! ++ ------------------------------------
On Tue, 25 Apr 2000, David Diaz wrote:
2nd I think this traffic balance issue is ridiculous and always have. If you have eyeballs they are paying for connectivity to the net, specifically content. In the rest of telecom those that initiate the request get billed so here eyeballs should be billed... and they are by their ISPs. If the network infrastructure is so expensive to maintain that the eyeball ISP losses money then they changed the incorrect rate. 1 word here caching. If the eyeball backbone wants to save network capacity and costs dropping some caches and give Doug H a call and add Cidera feed. If your network design is such that you cannot provide this caching service you better plan for it on your next build. The obligation of the big content backbones/websites should be to have multiple sites or inverse caching at the exchange cities to eliminate an undo burden on the eyeball backbones.
or why dont you require both to peer in 4 - 6 places in the US and honor meds both ways. This way, each is carrying traffic on their network. Of course the bigger providers dont want this as they would lose losts of $$ from all the payments they get for 'transit'. like i have said in the past, i dont think the DOJ or soon the EU is going to be happy once they find out what really goes on with peering... Christian
participants (7)
-
Christian Nielsen -
David Diaz -
I Am Not An Isp -
Jeff Barrows -
Larry Snyder -
Peter Galbavy -
Randy Bush