February 1997 - NANOG - lists.nanog.org

Directions to Grand Hyatt (?) [administrivia]
by Paul Ferguson 03 Feb '97

03 Feb '97

For those like myself who found the directions to the NANOG meeting at the San Francisco Grand Hyatt less than satisfactory (like, take the shuttle) I found satisfactory information by going to Map Quest [http://www.mapquest.com] ObSuggestion: It might also be a good idea to actually make directions available. - paul

2 1

Mailing list archive
by Eric Germann 02 Feb '97

02 Feb '97

Along the lines of the NANOG Search engine for meeting notes, the NANOG mailing list since the beginning of time, has been converted to HTML and indexed nightly at http://www.cctec.com/maillists/nanog/index.html Enjoy

1 0

Alpha test of MAE filtering capability
by Steve Feldman 02 Feb '97

02 Feb '97

In response to requests from several customers for a filtering solution at MAE EAST and MAE WEST, MFS WorldCom is developing an automated system to enable ISPs to control traffic destined for their own routers and networks. ISPs using this service will be able to specify, by IP address, which MAE-attached routers they wish to accept or block traffic from. On Monday, February 17th, MFS WorldCom will begin live alpha trials of this capability at MAE WEST with a small set of providers who have agreed to help us test this service option. We expect to extend the trial to MAE EAST shortly after and will notify you of the exact date. This means participating ISPs can begin filtering traffic to their routers at their discretion once the alpha period has started. Assuming the trial proceeds as expected, MFS WorldCom plans to introduce filtering as a new service option later in 1997. Please contact mae-info(a)mfst.com if you have any questions or concerns about the MAE filtering service. Steve Feldman

2 1

RE: Internet vs. Telephone company
by Paul J. Zawada 02 Feb '97

02 Feb '97

The Internet Thruway, IMHO, is the first serious attempt at getting to the root problem. I don't know much about Nortel's product, but I do know that using the PSTN seems to be the wrong way to get Internet service. Why on earth do you want to use a circuit-switched network to access a packet switched network? That's nuts, but unfortunately it's the only option Joe Consumer has. Cable modems? They'll be great when your cable company's plant will work birectionally and they get their data act together. I was amused by a letter to the editor in this week's _Fortune_ magazine. The writer proudly proclaimed that Internet access over cable worked wonderfully. He went on about how "..on the appointed day, at the appointed time, two cable installers, three Comcast personnel, and two CompUSA personnel arrived and successfully installed my cable modem in two hours..." Good God, if it takes 14 person-hours to install a cable modem, the cable companies are going to lose their shirts! (I'd like to know if this crew was even able to all show up in the same vehicle! :-) As for ADSL, I have serious doubts as to whether or not the phone campanies understand what to do with the technology. After they get xDSL local loops to everyone's home, will they know what to hook them up to? I've heard a lot about ADSL trials, but what are these loops being connected to? Does the solution scale? There are lots folks in the phone company who have no clue as to how the Internet works. This is why none of them saw this Internet access thing coming. (Not that I'm sure they could have done anything about it...) They now have a network that is underengineered for the way it's being used and they have to do something about it quick... The only means they know of is to raise the price. The Internet Thruway product has potential to help solve their problems, but I'm afraid that it may flop because it'll require too much cooperation between the telcos and the ISPs. After all, the telco nework was just fine until all of these ISPs showed up and screwed it up. You can be damn sure that most of the telco folks don't see it as thier problem to pay to solve this congestion in their network. Many still feel it's the telcos' God-given right to make a profit every quarter, regardless of how poorly they project future customer demand. I'm afraid that they're going to see Internet Thruway as an opportunity to charge ISPs for the privledge of helping to relieve the congestion in the PSTN... It all comes down to paying for more facilities, whether it's Internet Thruway, bigger switches and more trunks, or a parallel (ADSL-based perhaps) packet-switched network. The telcos won't make ubiquitous Internet connectivity happen for free. For that matter, nor would the ISPs if they were in the telcos' position... The fundamental problem is that until now, Internet users have been able to take advantage of "slack" resources in the PSTN at little or no additional cost. Now that these resources are drying up, no one (users, ISPs, telcos) wants to foot the bill for more. --zawada At 05:28 PM 1/31/97 -0800, Bharat Ranjan wrote: > >> My question is: Have other people heard of similar attempts by the >> telephone companies to have "metered Internet access"? >> It seems they would need a hardware level signal analyzer on their >> switching equipment to differentiate between voice/data. >>Does anyone know by what means such a technology could be implemented? > >Check out a product called 'Internet Thruway" that routes a call based >on the Called Number. The idea is for telcos to implement this product >as a back-end to their access network so calls being made to a Thruway >provider never reach the voice network. It is instead sent (via IP and >L2F) directly to the ISP's network. SBC is already committed to >deploying this in parts of Texas. Paul J. Zawada, RCDD | Senior Network Engineer zawada(a)ncsa.uiuc.edu | National Center for Supercomputing Applications +1 217 244 4728 | http://www.ncsa.uiuc.edu/People/zawada

2 1

Re: Internet vs. Telephone company
by Stephen Sprunk 02 Feb '97

02 Feb '97

>I think we're missing the point. They aren't just going to charge >hourly for data service, but for all service. > >It really wasn't that long ago they were doing this (US West >stopped in the late 70's or early 80's > >All they are going to need to do is notify customers and turn the >old billing procedures back on, they already track hourly usage >anyway. The Public Utility Commissions will have to approve it, which (in many areas at least) will never happen. Even if it did get approved, with local telco competition opening up now, there will undoubtedly always be some provider willing to offer unlimited access, regardless of what the Bells try. >>Tony Torzillo Stephen Sprunk

1 0

Internet vs. Telephone company
by Tony Torzillo 01 Feb '97

01 Feb '97

An interesting article was posted by the Seattle Times yesterday, it's at: http://www.seattletimes.com/topstories/browse/html97/inet_013097.html This was the Jan 30, 1997 issue. In it, Diedtra Henderson writes: "US West serves 2.2 million customers in Washington state and spent $340 million improving its phone system here last year, a spokesman said. The company blames a meteoric rise in Internet use for its phone-service woes. [SNIP] "For the next five years we are in big trouble. There is no clear, intelligent way to expand the phone system to handle the Internet demands," said Lu, who expects metered Internet access in the future - in other words, you pay by the hour. " My question is: Have other people heard of similar attempts by the telephone companies to have "metered Internet access"? It seems they would need a hardware level signal analyzer on their switching equipment to differentiate between voice/data. Does anyone know by what means such a technology could be implemented? What would be the legal ramifications? Also do people agree with the claims that our local lack of available phone lines is due to Internet usage or just to lack of foresight in growth management decisions? Tony Torzillo ---------------------------------------------------------------- Network Specialist, UW Network Operations Center, 543-5128 EMAIL: ndc-noc(a)cac.washington.edu, torzillo(a)cac.washington.edu ----------------------------------------------------------------

5 4

Re: Alpha test of MAE filtering capability
by Vince Fuller 01 Feb '97

01 Feb '97

Entirely agreed. On the other hand I have what is turning out to be a unique (here) point of view about this. I don't want to prevent this kind of theft -- I want to discover it, and remove perpetrators from any IXP where they try it. I don't want to block it. I want to ensure that it is never tried twice. I appear to be in the minority wrt this view. This is a great idea. But do you have any idea how much expert engineer time is involved in tracking down theft of this sort? Most of our employers would much rather spend that resource on tackling the big problems. If you can get the tools (simply having the IXP switch keep a port-to-port traffic matrix would help a lot...), I suspect you'd find a lot of converts to your "minority" view - many would prefer to expose and ostracize the perpatrators of these thefts instead of ignoring them. --Vince

3 2

Bogus route announcements
by Michael Dillon 01 Feb '97

01 Feb '97

This seems more appropriate here than on NAIPR. I took the liberty of removing the discussion that led up to it and leaving only Karl's words. This seems to tie in to the layer 2 filtering discussion here. ---------- Forwarded message ---------- Date: Fri, 31 Jan 1997 20:09:59 +73800 (CST) From: Karl Denninger <karl(a)Mcs.Net> To: Michael Dillon <michael(a)MEMRA.COM> Cc: naipr(a)lists.internic.net Subject: Re: Implied warranty of routability? Was: Re: US CODE: Title 15, ... [some discussion of bogus TLD's and bogus routes deleted] Balderdash. Just the other day 0.0.0.0/0 (yes, DEFAULT) was being propagated by a LARGE NUMBER of national providers -- from a rogue (and unintentional) announcement that came out of a particular firm in Virginia. This went on for well over SIX HOURS before it was stopped. It was transiting a large number of NATIONAL network provider's core hardware, and disrupting connectivity to a fair number of people, some of whom were completely clueless as to the cause. We found it because we run defaultless and ANY instance of default appearing in announcements or anywhere on our core is an instant five-alarm fire. When we finally called the guilty party (after informing peers and upstream links hours before with no effect), they had not heard ANYTHING about it as of yet, and the announcement was ALREADY a few hours old in our tables at that point. Filtered out quickly my tailfeathers. 99% of the companies out there don't filter ANYTHING at that kind of level. Try to maintain the filters on CISCO hardware to actually verify and prevent any rogue announcements -- good luck. You just can't do an EFFECTIVE job of this; the coordination you NEED to do so is completely non-existant between firms to make it possible, especially in the "swamp". Now you can get routes from only a route server, yes, and that does help. Quite a bit. But basically all providers of any significance have exchange point(s) where the RADB isn't used. If the address isn't something that someone else is using, and is of sufficient prefix size (in 206 and above) I bet it wouldn't be noticed for months -- if ever -- until someone tried to get a so-called "official" allocation of the same number and said "what the hell??" when they found it already in the tables. I bet I could announce a random "reserved" prefix and nobody would catch it for at least 30 days -- during which time it would work perfectly, and globally. Yes, doing that kind of thing would be highly antisocial. But don't think for an instant that anyone actually watches constructively for this kind of chicanery on the net. That would be a false assumption, as I think the little episode of the other day proves rather conclusively. -- -- Karl Denninger (karl(a)MCS.Net)| MCSNet - The Finest Internet Connectivity http://www.mcs.net/~karl | T1's from $600 monthly to FULL DS-3 Service | 99 Analog numbers, 77 ISDN, Web servers $75/mo Voice: [+1 312 803-MCS1 x219]| Email to "info(a)mcs.net" WWW: http://www.mcs.net/ Fax: [+1 312 248-9865] | 2 FULL DS-3 Internet links; 400Mbps B/W Internal

1 0

Re: Internet vs. Telephone company vs. cable company
by Tony Torzillo 01 Feb '97

01 Feb '97

I've wondered if the recent advent of cable modems could change our entire Internet/telecommunications strategy. Provided that the cable network becomes widely available, more internet users would shift from using telephone lines for access to cable lines. Couple that with the use of the internet phone, and the phone line would only be necessary for communicating with non-internauts. I can see a number of ramifications for this - companies making money from long distance usage would be rather upset - and backbone providers would have to upgrade existing circuits to handle the increase in traffic. Anyone have thoughts/comments in regard to this as well? Tony Torzillo ---------------------------------------------------------------- Network Specialist, UW Network Operations Center, 543-5128 EMAIL: ndc-noc(a)cac.washington.edu, torzillo(a)cac.washington.edu ---------------------------------------------------------------- On Fri, 31 Jan 1997, David Schwartz wrote: More than anything, the problem is due to a flat-rate monthly pricing plan for calls which are increasing in both number and length. David Schwartz On Fri, 31 Jan 1997, Tony Torzillo wrote: > My question is: Have other people heard of similar attempts by the > telephone companies to have "metered Internet access"? It seems they > would need a hardware level signal analyzer on their switching equipment > to differentiate between voice/data. Does anyone know by what means such > a technology could be implemented? What would be the legal ramifications? > > Also do people agree with the claims that our local lack of available > phone lines is due to Internet usage or just to lack of foresight in > growth management decisions?

3 2

RE: Alpha test of MAE filtering capability
by Chris A. Icide 01 Feb '97

01 Feb '97

---------- From: Paul A Vixie[SMTP:paul@vix.com] But let me turn it around. With no means of detection, why do we suspect that it's a problem? That is, why doesn't the cause for suspicion also work as a means of detection? Well, here is the way I found mine. We keep usage information on all of our router ports, and one day, my FDDI interface to an exchange point jumps by 10Mbps. I haven't added any customers, and going back to examine my traffic patterns for customer ports, I have no cooinciding traffic increase. However, I do show this increase mainly passing from one Exchange point to the other. After isolation all traffic sources that would have created such a jump in traffic, I come up with a big goose egg. So, my next step was to log some flows from the router at the exchange point, and after pouring through quite a few flows, I begin to see traffic from an entity that my company has absolutely no relationship with. This all takes quite a bit of time. I would not want to judge anyone with partial data. Meanwhile bandwidth paid for by my customers, and engineered based upon my customer's needs is being chewed up. My customers are affected. I would prefer to prevent such events from affecting my customers, who I think would agree with this method. IMHO, as long as money is involved, and as long as someone thinks that they have a chance of getting away with something, they will try it. Chris

7 6