- Test - lists.nanog.org

Re: Sprint Peering Requirements? WAS More network problems (fwd)
by Sean Donelan 16 Mar '96

16 Mar '96

>> In the meantime just dump the Sprintlink traffic through Sprint's T1 >> line at the CIX, and let Sprint backhaul it to the other coast. > >That link has been down for 19 hours. I'm not sure what the problem is; >the T1 IXC won't talk to me about it since it's not my circuit. Awsome, the sales people will love that :-). A policy against the use of available alternate paths results in a single failure turning into a prolonged outage. I love big providers, they make it so easy. At least Sprint has a policy. MCI's and AGIS's lawyers have been working on a policy for months. But haven't come up with a coherent one yet. P.S. what is the deal with the Santa Clara POP anyway? I've seen gas main explosions, circuits cut by a blowtorch (gee? maybe they're related), "missing" fiber, "lost" circuits, and other shenangans. This time I'm glad we followed WorldCom's (WilTel, whatevertheirnameisthisweek) site recommendation and used DC power for our equipment. But this POP is like working out of a telecommunications war zone. -- Sean Donelan, Data Research Associates, Inc, St. Louis, MO Affiliation given for identification not representation

3 2

RE: Allocation of IP Addresses
by Jim Fleming 16 Mar '96

16 Mar '96

Yes...and we are headed for some big time Internet crashes... Now that big bucks and lawyers are entering the scene, a new picture will emerge...those people will bring politicians and their constituencies will follow... This is the first "Election Year" in the U.S. where the Internet will play a major role in the outcome... ...maybe by the year 2000 there will be an Internet Party and candidate... Jim Fleming Naperville, IL ---------- From: Paul Ferguson[SMTP:pferguso@cisco.com] Sent: Friday, March 15, 1996 2:59 PM To: Jim Fleming Cc: 'NANOG List' Subject: RE: Allocation of IP Addresses At 01:58 PM 3/15/96 -0600, Jim Fleming wrote: > >The growth of the railroads in the United States is also a good story >to study. There are many parallels with the Internet story. > With the number of train crashes in the past couple of months, this comparison leaves a lot to be desired, Jim. :-) - paul

1 0

Re: Sprint Peering Requirements? WAS More network problems (fwd)
by Sean Donelan 16 Mar '96

16 Mar '96

>I got this from one of our downstream customers. > >Is there any truth to this ? >> Okay, here's what you can do...Have your ISP contact Sprint and >> complain that they don't have connectivity to CRL network >> Services. CRL has T3s at several places that Sprint is at, but Sprint >> won't peer with them (Sprint has some stupid new requirements, >> apparantly, requiring peering at 3 NAPs.) No problem - all of CRLs are >> on 7507s :-) >> >> Anyway, CRL is at 2 NAPs - Washington DC and PacBell. Sprint's just >> being stubborn. ...anyway, have them complain to Sprint. The contact >> person at CRL is Eric Mehlhaff (415) 837 5300, he's the cheif network >> engineer. Sprint should take CRL up on CRL's offer :-). CRL has been telling me for months CRL can't peer at the Washington DC MAE-East because all of CRL's BGP peer slots at Mae-East are on "allocation." And therefore CRL doesn't have room for any more peers in DC. mae-east> show cdp neigh detail [...] Device ID: gwdca1.crl.com Entry address(es): IP address: 192.41.177.104 Platform: cisco 2500, Capabilities: Router Trans-Bridge Interface: Fddi0/0, Port ID (outgoing port): Ethernet1 Holdtime : 166 sec Version : Cisco Internetwork Operating System Software IOS (tm) 3000 Software (IGS-I-L), Version 11.0(2), RELEASE SOFTWARE (fc1) Copyright (c) 1986-1995 by cisco Systems, Inc. Compiled Mon 25-Sep-95 18:34 by deannaw In the meantime just dump the Sprintlink traffic through Sprint's T1 line at the CIX, and let Sprint backhaul it to the other coast. Its a great sales opportunity when Sprint customers call ask why performance is so bad. Our network engineers may not like Sean Doran, but our sales people sure do. We've had several Sprint customers buy direct circuits to our network because Sprint's peering policies caused their traffic to travel the long way (and sometimes get lost along the way). As long as Sprint keeps refusing, you get to look like the helpfull company trying to meet the customer's needs. -- Sean Donelan, Data Research Associates, Inc, St. Louis, MO Affiliation given for identification not representation

2 1

Sprint Peering Requirements? WAS More network problems (fwd)
by Forrest W. Christian 15 Mar '96

15 Mar '96

I got this from one of our downstream customers. Is there any truth to this ? -forrestc(a)imach.com ---------- Forwarded message ---------- Date: Fri, 15 Mar 1996 11:07:57 -0700 From: Nate Williams <nate(a)sri.MT.net> To: forrestc(a)mt.net Subject: More network problems OK, here's the scoop, from the network weenie at Walnut Creek. > Okay, here's what you can do...Have your ISP contact Sprint and > complain that they don't have connectivity to CRL network > Services. CRL has T3s at several places that Sprint is at, but Sprint > won't peer with them (Sprint has some stupid new requirements, > apparantly, requiring peering at 3 NAPs.) No problem - all of CRLs are > on 7507s :-) > > Anyway, CRL is at 2 NAPs - Washington DC and PacBell. Sprint's just > being stubborn. ...anyway, have them complain to Sprint. The contact > person at CRL is Eric Mehlhaff (415) 837 5300, he's the cheif network > engineer. > > (pass this info on to Sprint, but don't have your ISP call CRL - Eric > would just be annoyed). > > Eric is trying to force Sprint to peer with them at DS3 speeds > (45mbit), and this might help. Right now, I can get to WC in-directly by going through sri.com, but this isn't acceptable as I need direct IP connectivity to get the latest source bits. Nate

1 0

Re: Allocation of IP Addresses
by Robert Moskowitz 15 Mar '96

15 Mar '96

At 02:20 PM 3/14/96 -0500, @NANOG-LIST wrote: >Brian asks (rhetorically?): > >> Show me how I can create more ip numbers today. >> > >(1) Create new numbers, call it IPv7; > IPv7 was the CATNIP proposal. I think you have to go out to IPv9 to get an unused version #. Robert Moskowitz Chrysler Corporation (810) 758-8212

2 1

RE: Allocation of IP Addresses
by Jim Fleming 15 Mar '96

15 Mar '96

The growth of the railroads in the United States is also a good story to study. There are many parallels with the Internet story. Another story to study is the telephone industry... In the "old days", telephone systems were based on a model where the central core switching system was very expensive and the telephones were cheap. Centralized authorities made some sense, partly because the equipment capitalization was enormous. Common control was easier and could be performed by a few expensive systems. With the Internet, the capitalization is largely incremental. The core "switching system" is low-cost compared to telephone systems. Unlike low-cost telephones, PCs and servers are placed on the "edges" of the system and help to contribute to the common control. This incremental capitalization allows people to start small and to grow at their own pace. Also, companies can select prime urban markets and bypass rural areas. This is similar to what happened with railroads. As companies grow (like railroads) they start to encounter other growing companies and they best serve their customers by coordinating routes that allow packets (passengers) to transfer from one place to another with ease. It does not do a customer any good to be stranded in Peoria with no route to Des Moines. The Internet is only one component of the rapidly evolving story of a decentralized telecommunications system. The Internet would not be here if UNIX had not been set free, if AT&T had not been split, and if PC technology had not been allowed to evolve without centralized controls. Unfortunately, the lack of centralized control allows the Internet to evolve but also discourages coordination. Eventually, customers (passengers) will be the ones that get to decide which ISPs (trains) they ride. ISPs that work together to provide customers with the best service will survive. Those that do not will end up as carnival rides without connections to anything. Jim Fleming Naperville, IL ---------- From: Kevin Oberman[SMTP:oberman@nersc.gov] Sent: Friday, March 15, 1996 10:08 AM To: perry(a)piermont.com Cc: Barry Shein; 'NANOG List' Subject: Re: Allocation of IP Addresses > Date: Fri, 15 Mar 1996 00:40:24 -0500 > From: "Perry E. Metzger" <perry(a)piermont.com> > > I'm arguing that they haven't ever really existed, except with > government intervention to create them. Guess this you could claim "government intervention" for most anything, but I suggest you read up on Carnagie Steel, Standard Oil, and a few others at the turn of the century. And, if you want to claim that these were local, go out and buy a diamond. While there are a few exceptions, DuBeers controlled the world's diamond market totally until the past decade and still could easily destroy the worlds diamond market in a day. Yes, monopolies have and continue to exist. (A monopoly on IPv4 address space is very improbable and would certainly be short-lived, though!) R. Kevin Oberman Energy Sciences Network (ESnet) National Energy Research Supercomputer Center (NERSC) EMAIL: oberman(a)es.net Phone: +1 510 422-6955

2 1

RE: Allocation of IP Addresses
by Jim Browning 15 Mar '96

15 Mar '96

>From: Perry E. Metzger[SMTP:perry@piermont.com] > >Jim Browning writes: >> DeBeers has controlled (or at least significantly influenced) the price of >> diamonds for decades, ensuring that they appreciate in value by controlling >> supply, and by buying diamonds on the wholesale market to keep the price up >> (similar to what countries do to protect their currencies). They could >> flood the market at any time, driving the price down dramatically. Of >> course this isn't in their best interest, so they won't do it. However >> Russia might, as they have stockpiled large quantities of diamonds, and may >> not have the same goal as DeBeers... > >So in what way is this a demonstration of a monopoly able to >arbitrarily raise prices? What you've shown is that someone might be >able to arbitrarily drop prices... The price has been artificially raised through the years by controlling supply, and it is only a government (the former USSR) which has (to the best of my knowledge) threatened DeBeers control. -- Jim Browning

1 0

RE: Allocation of IP Addresses
by Paul Ferguson 15 Mar '96

15 Mar '96

At 01:58 PM 3/15/96 -0600, Jim Fleming wrote: > >The growth of the railroads in the United States is also a good story >to study. There are many parallels with the Internet story. > With the number of train crashes in the past couple of months, this comparison leaves a lot to be desired, Jim. :-) - paul

1 0

RE: Allocation of IP Addresses
by Jim Browning 15 Mar '96

15 Mar '96

>From: Perry E. Metzger[SMTP:perry@piermont.com] > >"Kevin Oberman" writes: > >> And, if you want to claim that these were local, go out and buy a >> diamond. While there are a few exceptions, DuBeers controlled the >> world's diamond market totally until the past decade and still could >> easily destroy the worlds diamond market in a day. > >How could they "destroy the market"? Is there a "diamond market" sitting >in a hallway somewhere that I can shoot or something? DeBeers has controlled (or at least significantly influenced) the price of diamonds for decades, ensuring that they appreciate in value by controlling supply, and by buying diamonds on the wholesale market to keep the price up (similar to what countries do to protect their currencies). They could flood the market at any time, driving the price down dramatically. Of course this isn't in their best interest, so they won't do it. However Russia might, as they have stockpiled large quantities of diamonds, and may not have the same goal as DeBeers... -- Jim Browning

2 1

Re: Interesting AP Article
by HAKNER JEFF 15 Mar '96

15 Mar '96

As fascinating as these recent hot debates on YA Death-of-the-Internet scenarios, free markets, theoretical bounds on the food supply and an abundance of hoarded water have been......... What has been conspicuously absent is any comment from anyone at sprintlink on the original posting ("Interesting AP Article"). I tend to distrust anything I read in the popular press about tech topics. What is the truth level in that article's statement that sprint was "blocking traffic"?

2 1