June 1997 - Test - lists.nanog.org

nanog digest now available
by Allan Chong 01 Jun '97

01 Jun '97

Since I couldn't find a partner to do E.Buttress of Middle Cathedral in Yosemite this weekend, I wasted a couple of hours and installed a nanog digest on a friend's server. Right now, normal nanog mail will go to my account on the machine and then through procmail (which I've never really used before). I've hacked out a perl script to format up a digest with subjects at the top and cron it out at noon and say, 6pm. Users will be added manually by mailing me at: allan(a)alum.mit.edu I will setup the procmail as needed. Please email me your working procmail scripts to filter your favorite net kooks^H^H^H^H^Hpersonalities. I think the rule will be, "if you annoy me or another digest member, you're filtered." Later, I might add something to show the messages that got deleted, so everyone can know what they're missing. Yes, I know I could have done it with majordomo, but this just uses a normal account on the machine without having to play with their mail setup. I'm currently waiting on the subscription approval for the bogus user. allan PS. The digest looks like: This unofficial procmail'ed digest sponsored by Allan Chong. Please mail subscribe/unsubscribe requests to allan(a)alum.mit.edu today's topics: test12 (Allan Chong <allan(a)alum.mit.edu>) test13 (Allan Chong <allan(a)alum.mit.edu>) test14 (Allan Chong <allan(a)alum.mit.edu>) test15 (Allan Chong <allan(a)alum.mit.edu>) test10 (Allan Chong <allan(a)alum.mit.edu>) ##################################################### ##################################################### From: Allan Chong <allan(a)alum.mit.edu> Subject: test12 test12 ##################################################### From: Allan Chong <allan(a)alum.mit.edu> Subject: test13 test 13

1 0

[Fwd: unsubscribe]
by Adnan.Khan＠ds.emirates.net.ae 01 Jun '97

01 Jun '97

Please unsubscribe me

1 0

Re: connectivity outside the US
by mansboury＠yesvirtual.com 01 Jun '97

01 Jun '97

What you are describing is reasonable in a truly open market environment. However, the cost of committing to $650K to $800K for E1s limits the field significantly. Especially when the carriers bid on a large portion of the available bandwidth. The availability of 100G loops will be diminished by the more financially sound carriers (i.e., Telstra). I think you know those guys. When you look at the demand between now and the year 2000 and you guys will be looking at consuming +200M of the bandwidth for IP services and God knows how much for the more traditional services. Even though the investment has remained relatively constant the market pricing has not. I don't see it being a game for the faint hearted. And the massive consumption from the likes of AT&T, BT/MCI and Worldcom will greatly influence the pricing. You may see a trend towards reducing the price for the bulk buyers/investors, but it will not likely come down as fast as your predict for the low to moderate users. Your assessment may be right, but I am not so sure we will see it before the year 2000. Mark ---------- > From: Geoff Huston <gih(a)telstra.net> > To: Mark Ansboury <mansboury(a)yesvirtual.com> > Cc: Sean M. Doran <smd(a)clock.org>; avg(a)pluris.com; bgreene(a)cisco.com; jesse(a)netthink.com; map(a)iphil.net; nanog(a)merit.edu; smd(a)clock.org > Subject: Re: connectivity outside the US > Date: Sunday, June 01, 1997 6:04 PM > > Supply and demand - the cables which are being commissioned in late 98 and > 99 look to be 100G loop designs in bot the Pacific and Atlantic > environments. The system price looks like remaining pretty constant - > around $500M give or take a few hundred mil - despite a 50 fold increase in > capacity. Its reasonable to expect a radical bottoming out of unit price > for IRUs sometime in 99 for these cables. What does that mean? The scarcity > premium in cable price which will be evident through 98 will be eliminated > in 99 and insterad you'll see a return to a technology-indexed unit price, > where the price will reflect the basic project costs And given the increase > in cable capacity due to WDM deployment you can expect a unit price drop. > > > > At 12:18 PM 1/6/97 +1000, Mark Ansboury wrote: > The cost for > >overseas cables are not going to be coming down in a big way over the next > >couple of years. They will more likely go up. At least for our foreseeable > >future. The next few years. The pent up demand is to great. >

1 0

Re: connectivity outside the US
by Hal Murray 01 Jun '97

01 Jun '97

> A few months ago a friend of mine from Fore Systems told me about > the 'ACTS' project. NASA and Ohio-State and quite a few others were > doing research into the Gigabit Satelite Network. > Anybody heard anything about this? The Advanced Communications Technology Satellite (ACTS) provides for the development and flight test of high-risk, advanced communications satellite technology. http://kronos.lerc.nasa.gov/acts/acts.html

1 0

Re: connectivity outside the US
by mansboury＠yesvirtual.com 01 Jun '97

01 Jun '97

The bottom line throughout this discussion is that neither the techies or the capacity planners really make the divisions. The economics of running cable is only a small factor. It goes to supply and demand. Scarcity creates higher pricing. Look at the recent bids on the new cable??? Everyone wants a piece and some carriers would like to monopolize as much bandwidth as possible to control the flow. The accountants and the marketing types look to optimize their return early in the game. What does that mean for the Internet? The overseas ISPs are building up their resources. The only ISPs that can really afford to stay in the game in a big way "over the long haul" are now the carriers themselves. The end result??? The marketing and pricing strategies of the ISP and market are being commercialized and in the overseas markets looking more like the pricing strategies of the typical telephone service offering.The cost for overseas cables are not going to be coming down in a big way over the next couple of years. They will more likely go up. At least for our foreseeable future. The next few years. The pent up demand is to great. My humble opinion. Mark T. Ansboury Virtual Resources Inc. Virtual Communications Inc. E-mail: mark.ansboury(a)yesvirtual.com

2 1

Re: connectivity outside the US
by Vadim Antonov 01 Jun '97

01 Jun '97

smd(a)clock.org (Sean M. Doran) wrote: > Examining this a bit more closely, since undersea capacity is > terribly expensive, when there is adequate capacity available > to a large aggregate of sites people want to get to, there will > be an obvious market for access to that capacity. Actually, i do not understand why undersea capacity is so expensive. Cable is more expensive, yes; but the paths are much straighter, and there's no need to purchase rights of ways (except for shore-side strips). There's no need to dig trenches -- you just drop the cable off the boat. I guess the real problem with undersea capacity is more in the fact that it was always considered a low-volume service (which it is, in terms of voice traffic); so there's no many competitive providers, and small-quantity pricing. --vadim

6 5

NANOG Host for October
by William B. Norton 01 Jun '97

01 Jun '97

Hi - Thanks to those that have volunteered for hosting NANOG meetings in the past. We now have an opening for the October '97 NANOG 11 slot. If you are interested in hosting this NANOG, please read the "Tips for Hosting a NANOG" (http://www.nanog.org/tips.html) document on the www.nanog.org web page and send me e-mail. Thanks! Bill -------------------------------------------------------------- William B. Norton <wbn(a)merit.edu> (313) 764-9430

1 0

Re: RFC 1918 addresses
by Joseph C. Pistritto 01 Jun '97

01 Jun '97

While this is technically correct, actually changing the addresses in ICMP payloads violates my first rule of debugging complex systems: Diagnostics should be a simple as possible and never be altered by any non-diagnostic system. (Which begs the question of should people be using these addresses locally at all. Which is kind of metaphysical except that it *does* preserve address space, which is a universal good. Some people do this for security reasons too, although it's at best only moderately effective security measure and could produce a false sense of security inside such an addressing realm.) I agree that ever having a source or destination IP that's RFC1918 outside the domain is a very bad thing. -jcp- ---------- > From: Paul A Vixie <paul(a)vix.com> > To: nanog(a)merit.edu > Subject: RFC 1918 addresses > Date: Saturday, May 31, 1997 4:38 PM > > I think that any exposure of these addresses outside their addressing realm > is a mistake. Using them for otherwise unnumbered serial links would be fine > if Cisco had an option to "use loopback address when sending ICMP's" but they > don't. Traceroute is sending increasing-TTL'd UDP datagrams, and if you are > seeing a 10.0.0.2 source address on an router's ICMP to you it means you would > get that as a normal ICMP too -- meaning not just one due to a diagnostic like > Traceroute. I think this is an error. > > Exposing an RFC 1918 private address in, say, a "Received:" header in e-mail > is less of a problem, though the spammers who do it are actually better able > to cover their origins, there's no way to prevent it and no normal damage > from doing it. > > No IP datagram with an RFC 1918 address in the protocol headers should be > allowed outside a private addressing realm. This means not the source > address, not the destination address, and not the encapsulated addresses > inside an ICMP payload.

2 1

Re: connectivity outside the US
by smd＠clock.org 01 Jun '97

01 Jun '97

"Miguel A.L. Paraz" <map(a)iphil.net> writes: > There can be two ways of looking at it, from the WWW/content point of > view: > > * "They're unlucky enough to be outside North America, but they want > American content anyway. So they have to suffer with poor > performance and high prices." > * "These folks outside America don't get to appreciate our content. > Their numbers are growing. Thus we have to invest in the delivery > infrastructure." Of course that leads to: * We have to worry about international copyright issues as overseas ISPs begin duplicating our content to save on IPL bandwidth * We should deploy duplicate servers in other parts of the world so that we don't have to deal with the problems of shortages of IPL bandwidth On the other hand I argue that the fundamentally interesting thing about the Internet is that traffic patterns shift substantially at little notice through the development and deployment of new technology. Moreover, in in Internet whose origins were "every client is also a server" and where, with proxying and the like, that is still essentially practical, a strict content:consumer mindset is a very short-sighted business approach. > But, the Independent Big Network people will still have to get > their own IPLs - which may end up with BT/MCI, Sprint/Global One, > or AT&T? What's the difference between becoming an IP customer > and just a leased line customer? Independent Big Network people are working on a simple risk statement: traditional bandwidth-owning telcos are too slow to keep up in the Internet marketplace, particularly when they're distracted by what various bits of deregulation are doing to their traditional income, creating an opportunity for the Independents either to thrive on their own or to be strong acquisition targets. Moreover, one might bet on partnerships with organizations competing with established telcos on other fronts which are becoming more like commodity products (most notably raw bandwidth). > How do we come up with a settlement model? Ultimately this depends on what everyone's real costs are. Alot of what is perceived as "cost" is due to misapplication of technology. A settlement model that encourages all parties to do the right thing to sustain the industry's growth is really what's needed. The problem is getting such a model agreed to and put into effect other than bilaterally among clever operators. Hahaha, good luck if there are American lawyers involved. > And more providers should be hauling in their own high-capacity pipes and > putting up their own overseas POPs instead of having the local folks > drag their lines to the US, and charge a lot for slow > connections. It's coming. The problem is that the people most likely able to do this are also the people earning half-circuit revenues from the large numbers of IPLs that would get aggregated into an IPL where both half-circuits would be paid by the provider entity. This is especially touchy when money for one of the half circuits is paid to another telco. In this way large capacity-owning telcos are in a bind about competing with a lucrative revenue stream (IPLs) whenever they do this and also when they compete against non-telco Independent Big Networks who might buy IPLs at market value. To make up for lost revenue (and maybe real cost) from the internally-deployed IPL, lots of traffic would have to be aggregated. This kind of massive overselling of capacity is still somewhat alien to the managed services mindset and even if it didn't it doesn't make it all that much easier for them to compete on "quality of service" profitably just now without some magic extra-high-quality pricing bullet. Sean.

1 0

RE: connectivity outside the US
by Sean M. Doran 01 Jun '97

01 Jun '97

Barry has an excellent point, and I'm glad that someone with his, um, background is helping undersea cable operators and Internet operators understand each other better. The not very extensive history of private undersea ventures has revolved around "cream-skimming" -- taking high-revenue high-capacity traffic from consortium-built cables -- and bandwidth appreciation. Essentially the private cable operators want to walk a line between having to depreciate their assets due to lack of revenue and having enough spare capacity for those times when consortium partners run short. That is, if you have a private cable, you don't want to sell all your capacity to end users because you are hoping that your competitors will run short on bandwidth, allowing you to jack up your prices. If and when you sell capacity you want it in long-term contracts, preferably for restoration of other paths so that if someone builds a competing cable or the accountants start demanding some return on investment you can do short-term "as-available" deals with government-funded research networks. Unfortunately this model does not tend to drive down prices substantially for most users, and particularly not for most users looking for an easy growth path (any kind of easily-tuned end-to-end TDM in theory should change this). Even turning unused capacity over for internal commercial purposes (and it is unclear whether UUNET would be an "internal purpose" given the joint ownership of the Gemini company) would be a huge risk as it involves tying capacity to a revenue stream, which means making hard decisions when someone wants to pay real money for that capacity when they run into a crunch. Terrestrial capacity politics have some similarities, btw, however the acquisition of rights-of-way in some places is so difficult that alot of the cross-leasing games revolve around trading capacity along various paths as much as playing the you're out of bandwidth pay us lots of money games (which AT&T does remarkably well, btw, in the U.S.). What becomes interesting is alot of new rights-of-way opening up and making it possible to offer dark fibre along them. This directly threatens the price escalation model as well as the managed-services pricing schemes bell-heads like. This is good for people building parts of the Internet. Sean.

1 0