- Test - lists.nanog.org

RE: Internet vs. Telephone company
by Robert Wilson 01 Feb '97

01 Feb '97

A couple observations: 1. Local and Long Distance phone companies stand to gain the most from the current Internet growth frenzy, yet seem to be the quickest to try to slow or defer its growth...They have the luxury of having enough a buaracracy surrounding them to play both ends AND the middle. One cannot say that they haven't profited from their growth....re-investment in infrastructure is simply par for the course when providing network services. 2. If they are NOT careful, TCI/@Home will have their lunch. Although ADSL may be viable, its cost model still looks high to the consumer when positioned against cable modems, and would offer no real tangible difference to the casual end user. My wish is that they begin to understand this at a macro level.... >---------- >From: Tony Torzillo[SMTP:torzillo@cac.washington.edu] >Sent: Friday, January 31, 1997 4:24 PM >To: nanog(a)merit.edu >Subject: Internet vs. Telephone company > >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 >---------------------------------------------------------------- > >

1 0

RE: Internet vs. Telephone company
by Bharat Ranjan 01 Feb '97

01 Feb '97

> 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. Check out the full story at www.nortel.com and search on 'Internet Thruway' Attached is a piece of the press release mm mm sssss nnnnnn * Bharat Ranjan * m m m s nnnnnnn * Network Engineer * m m m sssss nn nn * MSN Network Services * m m s nn nn * (206)-936-0471 * m m sssss nn nn * bharatr(a)microsoft.com * ******************************************************* * The opinions/ideas in this memo are not necessarily * * those of Microsoft Corp. * ******************************************************* ______________________ Nortel Announces Internet Thruway for Public Carriers Research Triangle Park, NC -- Nortel has announced Nortel Internet Thruway, the industry's first high-performance, end-to-end solution to help public carriers more efficiently handle the impact of rapidly growing Internet and telecommuting traffic. Available today, Nortel Internet Thruway works in a multivendor switching network to elegantly groom long-duration data and Internet calls off the voice network to a data network, minimizing the impact of the traffic on the voice network and allowing carriers to offer new services to data users, Internet Service Providers (ISPs), and companies offering remote access and telecommuting. Internet Thruway recognizes data calls based on called number and totally redirects them away from the originating switch before they reach the voice network. These calls are converted to packet data format, encapsulated in IP packets using industry-standard Layer-2 Forwarding (L2F) protocol for security, and sent to the appropriate ISP or corporate LAN over a frame relay or asynchronous transfer mode (ATM) network. Voice calls proceed to the voice network over a standard TR303 link, an open interface to central offices.

1 0

so are we talking product opportunity here?
by Paul A Vixie 01 Feb '97

01 Feb '97

An Alpha and NNSTAT and GateD running as part of a DAS ring that otherwise just has an NSP router and a GIGAswitch port, with a lot of hand tuning to notice "bad" flows, would definitely help discover "unintended transit" situations. Surely there's got to be someone on this list who wants to tackle the opportunity to build and sell such a box.

2 1

Re: Internet vs. Telephone company
by ! Drive 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. Who is to say that my data call is actually connecting to the Internet? >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? I would say an intential lack of foresight in order to increase charges for higher profits.

1 0

Re: Alpha test of MAE filtering capability
by Tony Li 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. Locks keep honest people honest. Tony

2 1

Re: Alpha test of MAE filtering capability
by salo＠msc.edu 31 Jan '97

31 Jan '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 sounds like a line of reasoning that would argue against filtering in routers: "I don't want to prevent [this activity, whatever it might be] -- I want to discover it..." What sort of filters do you put in your routers? -tjs

1 0

Blocking SPAM at the source?
by Michael Dillon 31 Jan '97

31 Jan '97

>From inet-access On Fri, 31 Jan 1997, Scott Hazen Mueller wrote: > One way to fix this would be for the sites that own large dialin pools to > firewall their dialins so that they can only do SMTP to their own servers. Is anyone doing this like IBM Advantis, InternetMCI, AT&T Worldnet, etc? If not, why not? Michael Dillon - Internet & ISP Consulting Memra Software Inc. - Fax: +1-250-546-3049 http://www.memra.com - E-mail: michael(a)memra.com

1 0

Re: "portability" of IP Addresses
by Eric Sobocinski 31 Jan '97

31 Jan '97

On Thu, 30 Jan 1997 at 21:09 MST, Pete Kruckenberg <pete(a)inquo.net> wrote: > > On Thu, 30 Jan 1997, Alan Hannan wrote: > > [...] > > Can a provider forbid their customer from announcing allocated > > networks to another provider? Yes. This is the discretion of the provider. Few forbid it completely, but it does happen. More likely you'll see a list of conditions that must be met before this is allowed, perhaps even including a small additional fee to cover the added administrative headaches. > > Assume the following situation: > > ------------------------------ > > > > ISP A gets customer C to sign an agreement for service. ISP A > > allocates network A.B.C.0/20 to customer C. A.B.C.0/20 is a > > subnet of CIDR block A.B.0.0/16, allocated to ISP A by our friends > > at the InterNIC as a non-portable address space. > [...] > > So, customer C goes to ISP B, and says, I'd like to buy service > > from you, and announce network A.B.C.0/20 to you. > > I would think that ISP A would be all for it, since as soon as ISP B > starts announcing the /20, all in-bound traffic will come through ISP B > and relieve some of ISP A's bandwidth. Since ISP A is announcing a /16, > the /20 announced by ISP B would take precedence, being a more specific > route. Why exactly do you think this makes ISP A happy? Most likely the link to ISP B is a link that ISP A would have preferred to sell themselves. And to "relieve some of ISP A's bandwidth" doesn't make ISP A any happier since they have to engineer for that bandwidth anyway for those cases where ISP B fails. Ok, I'll allow that some ISPs get charged by the packet by their upstream provider and thus would have slightly lower overall costs as ISP A under this scenario. I'd be surprised if that difference offsets added costs elsewhere in the average case. > I would think that ISP A would not have much to stand on from a legal > standpoint: > - your contract does not prevent you from buying service from another > provider (I'm assuming this), Probably true, but not necessarily so. There's nothing illegal about an exclusive contract. > and (I further assume) it does not > prohibit you from letting another ISP announce those routes This is a stretch. Chances are that your contract does not specify anything about announcing routes to third parties, but since the larger block was initially assigned to ISP A, it reasonably follows that ISP A can regulate those blocks through separate policy so long as your contract does not say otherwise. Even so, many (most?) contracts these days make sure by proclaiming that addresses assigned by the ISP are not the property of the customer and that the ISP retains some level of control. > - as it stands now (and will until ARIN changes it), ISP A does not > actually *own* an IP addresses, and there is at least one RFC that > makes this pretty clear, as well as several InterNIC policies This is true, but neither do you the customer. > - neither you nor ISP B are prohibiting ISP A from fulfilling their > obligations to you or the rest of your customers by announcing the > /20 separately > - the announcement of the /20 does not cost ISP A anything extra, nor > does it increase any contractual requirements for ISP A In an ideal, well-informed world, these latter two would be true. But they're not. The number of possible problems with external routes multiplies with multiple announcements. Then, when those problems do occur, ISP A and ISP B must both work at fixing things. If ISP B is cooperative, coordination still requires extra time and effort. If ISP B is not cooperative, ISP A either spends much more time trying to troubleshoot ISP B's part, or ISP A gets a black eye when ISP B screws up, or both. As more ISPs learn this, more are writing provisions for it into their contracts. > This is a pretty gray area, though, but I think you have a pretty good > chance to be able to contest what ISP A says. That probably won't make > them go out of their way for you in the future, though. That may or may not be true. An professionally-run ISP is happy to explain the logic behind their policies and work with you to find an adequate solution to your requirements. Questioning their initial ideas might yield better solutions, but forcing your solution over their objections is likely to yield a solution that doesn't fit your ISP's architecture and thus doesn't serve you well. And that WILL make them annoyed with you as well. > Pete Kruckenberg > pete(a)inquo.net Eric Sobocinski sobo(a)merit.net

3 2

Re: Alpha test of MAE filtering capability
by Vince Fuller 31 Jan '97

31 Jan '97

I think that this is the wrong approach. Better to monitor it, prove that it happened, and remove offenders from the IXP's altogether. The IXP contracts should include just such a provision. I don't disagree. Do you have tools for reliably detecting unacceptable default route usage? Until those tools are available and until IXP's agree to police their interconnects, the MAC filtering approach is the only real solution available. Your fellow IXP members are deserving of your trust, until they show that they aren't, and the paternalistic "let's remove the temptation" approach is just offensive. This is a nice thought, entirely approprate for a more cooperative Internet of an earlier age. Unfortunately, today it would seem that "better fences make better neighbors" is a far more practical philosophy. --Vince

2 1

LAST CALL
by Pam Ciesla 31 Jan '97

31 Jan '97

For the February 10 & 11 NANOG Meeting held in San Francisco. Last day for registration is Friday, January 31. To advoid delays at the meeting please send in form of payment ASAP. Thanks Pam Ciesla

2 1