- Test - lists.nanog.org

Re: Peerage versus Peering
by William Allen Simpson 05 May '97

05 May '97

I'll try not to belabor the point too much, since this fellow copied my previous message in its entirety in his reply (I do wish folks would learn how to use a MUA), but.... > From: "Jeff Young" <young(a)mci.net> > webster certainly never contemplated this form of 'peer' so > it is useless to quote him. i agree with peter, in this > form 'peer' means a network of equal or similar size. in > the current state of technology, peer to me means capable > of asymmetry. > The folks discussing peering and routing policy, lo these many years ago, were relying upon webster and other sources like unto it for the terms to use in describing our needs. For example, see RFC 1104. > i'm sure the rest of nanog will play a large role in defining > this term 'peer' in the coming months, native english speakers > and not. > If you want to define some term for "networks of equal or similar size", please use some other term, since the use of "peer" in network terminology is already taken. "Oligopoly" comes to mind. WSimpson(a)UMich.edu Key fingerprint = 17 40 5E 67 15 6F 31 26 DD 0D B9 9B 6A 15 2C 32 BSimpson(a)MorningStar.com Key fingerprint = 2E 07 23 03 C5 62 70 D3 59 B1 4F 5E 1D C2 C1 A2

1 0

RE: UUNET Pulling Peering - Core Issues
by Hong Chen 05 May '97

05 May '97

I think that what UUNet just announced has far more impact than just forcing smaller ISPs to pay for the peering. Essentially, the umlimited flat rate service offered by smaller ISPs will be gone, and replaced by metered services. That is, you pay what you use. Because of UUnet and other large ISPs' demand for peering ($60K/month for DS3?), the operating cost to regional ISPs will go up. The regional ISPs will have to raise their fee. There are other problems that will come soon after UUNet's announcement. That is the issue of 'who should pay for the international lines'. International lines are quite expensive and I remember that Tokyo Internet just paid US $20M for a 4 year DS3 contract from Japan to US. According to many major ISPs overseas, today's US Internet Centric status will not last long when Internet grows tremendously overseas, and some day, US national ISPs either have to build their own International links, or they have to share the international link cost with the oversea providers, just like today's telecom model. I can even see that global ISPs such as UUnet, MCI, etc one day will charge large national ISPs in the US for peering fees as UUNet and MCI think that they have paid the expensive international links. What does this mean? It means that peering fee will likely go up again (bandwith also grows), and cost to the regional ISPs can be even higher. The regional ISPs will have to ask themselves 'does it make sense to charge $19.95 for a customer while this customer is sending unlimited number of emails worldwide, and making unlimited internet phone calls worldwide?' To cover the cost, should the regional ISPs raise the fee to $30/month for unlimited? Maybe a tired fee structure will make sense, like what Netcom is doing today. The question begs again: Does the smaller ISP or regional ISPs have the necessary software/etc to bill for the metering or tiered services .... Sounds like that a conference can be held for the afermath of 'ending of free peering' Hong Chen hchen(a)aimquest.com On Sat, 3 May 1997, Charles R. Hoynowski wrote: > > > Uh, excuse me? I have no right to count on complete Internet connectivity > from my ISP? This has some far reaching implications. > > What consideration is being given to arguably the most impacted parties in > this new peering (or not peering) policy? The business customers, selling > services via the Internet. Consider, company "A" on Alternet is contracted > to provide data on demand to support services running on company "B", who > is on another ISP. Without warning or notice to its customers (which may be > the most disturbing point) UUNET makes a policy decision to stop peering > with the other ISP. Company "B" goes out of business. "B" sues "A" for > breach of contract, "A" sues UUNET, etc. > > Should ISPs be allowed to set peering policy for Internet? Should they have > the right to define what the Internet is? > > Which points out how the Internet really has become a common carrier/utility > company concern. PacBell can't suddenly change your phone service because > they decide to. There's public review and approval steps they have to > follow first. And there's good reason for that. The public relies on the > phone system for business transaction and private use. The system must be > be maintained in a consistent manner to allow that to occur. I don't need > to point out the similar common carrier roles ISPs and the Internet as a > whole have assumed. > > The issue has brought to light the fact than players in the ISP game want > to assume the role, and make profits based on, of providing common utility > services, without assuming the responsibility and accountability structures > other public service providers must assume. > > I believe this is the core issue at hand regarding UUNET's peering policy > changes and needs to be address at that level. > > > _/_/_/_/ _/_/_/_/ _/_/ _/ _/ Charles R. Hoynowski, Site Admin > _/ _/_/ _/ _/ _/ _/ Etak, Inc., 1430 O'Brien Drive, > _/_/_/ _/_/ _/_/_/ _/_/ Menlo Park, CA 94025 > _/ _/_/ _/ _/ _/ _/ Voice:415-617-4458 Fax:617-0161 > _/_/_/_/ _/_/ _/ _/ _/ _/ Email: charles(a)etak.com > >

3 2

RE: set community no-export
by Steve Hines 05 May '97

05 May '97

Seems to be an 11.1(x) and beyond command - I find it on 11.1(8), but not where I'm still running 11.0(11). - steve h - >---------- >From: Craig A. Huegen[SMTP:c-huegen@quadrunner.com] >Sent: Monday, May 05, 1997 10:33 AM >To: Sean Donelan >Cc: nanog(a)merit.edu >Subject: Re: set community no-export > >shOn Sun, 4 May 1997, Sean Donelan wrote: > >==>It also seems to confuse the heck out of the support folks when they >==>need to troubleshoot things. It would really be nice if cisco's had >==>a "show ip bgp neigh xxx out" command that showed what you are telling >==>your neighbor. Essentially the inverse of the "show ip bgp neigh xxx >==>route" command. > >In later versions: > >IOS (tm) 4500 Software (C4500-I-M), Version 11.1(9), RELEASE SOFTWARE (fc1) > >li-rt-01#show ip bgp nei w.x.y.z ? > advertised-routes Display the routes advertised to an EBGP neighbor > >/cah > >

1 0

US Seeks Alternative Domain Name Plan
by Cameo Wood 05 May '97

05 May '97

by Rebecca Vesely 6:14pm 2.May.97.PDT The Clinton administration, dissatisfied with a new international agreement on the registration of domain names and other key Net issues, has directed a working group to amend the system, a government official told Wired News on Friday. The official, who is knowledgeable about the policy discussions and spoke on condition of anonymity, said that the US initiative is aimed at easing emerging disputes over the International Ad Hoc Committee's pact and creating a more tangible mechanism for handling far-reaching Net issues. "While we feel that the IAHC has done a good job, we are concerned about numerous aspects of their plan," the official said. A central problem, the official added, was the role assigned to the UN-affiliated International Telecommunications Union and World Intellectual Property Organization without approval from member states. Full text available at http://www.wired.com/news/politics/story/3603.html

1 0

set community no-export
by Sean Donelan 05 May '97

05 May '97

No that doesn't work very well, unless you make the rash assumption that all of Sprint's customers are in the same Autonomous System. We saw all sorts of goofy problems when I*STAR did this with MCI. The problems went away when the customer switched from I*STAR to UUNET/Canada :-) You need a community that will be announced to "customer" BGP sessions, but not to "peer" BGP sessions. Its not hard to do, but it does seem to be hard to explain to your standard order taker/sales staff. It also seems to confuse the heck out of the support folks when they need to troubleshoot things. It would really be nice if cisco's had a "show ip bgp neigh xxx out" command that showed what you are telling your neighbor. Essentially the inverse of the "show ip bgp neigh xxx route" command. >set community no-export? > >> It's the other direction that is the problem. How does he advertise >> routes to Sprint so that Sprint customers can reach him but without >> Sprint advertising his routes at exchange points? That's the other >> half of what he's asking for, and he can't do it unless Sprint sets >> up filters to do it for him. -- Sean Donelan, Data Research Associates, Inc, St. Louis, MO Affiliation given for identification not representation

3 2

Re: eDNS - Temporary Freeze
by Dave Crocker 05 May '97

05 May '97

At 2:12 AM -0700 5/5/97, Karl Denninger wrote: >Blackholing the routes would be, IMHO a serious legal issue. Since it is a technique already in significant use and gaining in popularity, its application to this context is unlikely to raise new legal questions. d/ ---------------------------- Dave Crocker, Director +1 408 246 8253 Internet Mail Consortium (f) +1 408 249 6205 127 Segre Place dcrocker(a)imc.org Santa Cruz, CA 95060 USA http://www.imc.org Also: iPOC member, expressing personal opinions http://www.iahc.org

1 0

Re: Peering/Interconnection on the Internet as a Telecom Carrier
by Stephen Balbach 05 May '97

05 May '97

> I am a proponent of some level of regulation in the Internet as soon as > possible. I think this minimal level of regulation is critical to the > Internets open and unrestricted development. David - Is not regulation and unrestricted development at odds? The communications Act of 1996 is an attempt to remove restrictions put in place in 1934. The history of the telephone is an excellent of example of why not to regulate. The 40-years after 1934 was a Tellecommunications dark-age. Little innovation, diversity of service or choice. The Internet industry is too young to keep stagnet and survive in a competitive world market. > As an entrepreneur and a professional in this industry, I urge some > minimal level of immediate regulation such that the application of > accepted telecommunications law and practices are upheld. Are the accepted telecommunications law and practices such a good thing to uphold? > If they are not, I fear we will suffer the consequences of the > accumulation of all the economic power over the Internet in to a very > small number of hands. To me this is a very big problem, perhaps one even > worth losing your job over. David, you stood up for what you believe in and I have a great amount of respect for that. Perhaps in a few years we can look back when UUNET and a few others have put all the other ISP's out of business, and wonder why somthing was not done. You have a major battle to fight to convince people of that fact, and if you can, the rest will fall in place. I personally do not agree that it will happen as I am a great believer in the free markets ability to self-heal. After all, its just people, and no one likes a company that doesnt play fair. .stb

1 0

RE: eDNS - Temporary Freeze
by Geoffrey Waigh 05 May '97

05 May '97

Karl Denninger [SMTP:karl@Mcs.Net] wrote around Sunday, May 04, 1997 12:57 PM (in some timezone, my mail client is useless): > >As long as the Internet's cache files on each and every system out there >point at "a" in their file, there is the potential to break the namespace. > >One rogue server in a confederation will cause serious problems. > >NSI has defacto control, because getting them out of the cache files is a >long and slow process, and until they ARE out their answers will be >believed. That's one of the nice things about widespread backing of the IAHC plan by the big players. It wouldn't be hard for them to blackhole the cached rogue servers and thus cut them out of the loop. No doubt Karl will invoke lawyers at this, but while people are running around seeking injunctions to support their own lawlessness, the net keeps running. I would think that network operators could argue this is no different from filtering out rogue routes coming in from a subverted BGP peer. Geoffrey

3 2

None
by Jeff Young 05 May '97

05 May '97

1 0

Re: DNRC Press Release (fwd)
by Janet Pippin 05 May '97

05 May '97

uhh.. ----- Forwarded message from KathrynKL(a)aol.com ----- Date: Sun, 4 May 1997 23:06:22 -0400 (EDT) To: hostmaster(a)cyberramp.net, kathryn(a)domain-name.org Subject: Re: DNRC Press Release (fwd) Thanks. We are indeed having server problems but they are being fixed. Please try www.domain-name.org tomorrow to see a variety of materials about our organization. Many thanks for your interest, Kathryn Kleiman President, Domain Name Rights Coalition << FYI- Your server www.domain-name.org appears to be down. I could not get to it through a net browser, and could not ping the site (see below). ----- End of forwarded message from KathrynKL(a)aol.com ----- -- Janet Pippin * CyberRamp Internet Services Network Administrator *** 11350 Hillguard Road jdp(a)cyberramp.net * Dallas, Texas 75243-8311 http://www.cyberramp.net * (214) 340-2020 (817) 226-2020 ----- End of forwarded message from jdp ----- -- Janet Pippin * CyberRamp Internet Services Network Administrator *** 11350 Hillguard Road jdp(a)cyberramp.net * Dallas, Texas 75243-8311 http://www.cyberramp.net * (214) 340-2020 (817) 226-2020

1 0