- NANOG - lists.nanog.org

NOTES now have a search engine
by sob＠academ.com 30 Jan '97

30 Jan '97

I have added a search engine on the NANOG notes. You can get there at this URL: http://www.academ.com/Architext/AT-NANOG_NOTESquery.html

1 0

What will they think of next?
by Steven Eric Rubin 30 Jan '97

30 Jan '97

I pulled this off of ba.internet, thought you might find this interesting.. Name of poster removed to changed to protect the less enlightend --- Hi folks, Many of you already know me. I've been around in ba.internet and ba.news for a long time, and I'm not one to spam. I've been looking at the problem of congestion for quite a while now, trying to come up with a reasonable plan - and I think I have. For the long answer, see [sorry, Im protecting the guy]. For the short answer - how about a place where the small ISP can come, locate a router, and peer with a bunch of other ISPs? You don't need PI space, you don't need BGP4, and you don't have a $3000 per month co-lo bill. We will peer using OSPF, which almost all routers support. This will keep the local traffic local, and allow the small ISP to actually get a reasonable price for bandwidth to other local sites.

2 1

RE: peering charges?
by Vadim Antonov 30 Jan '97

30 Jan '97

Eric D. Madison <madison(a)acsi.net> wrote: >Well, that's not exactly true, UUNet nor PSI show "profits", they show >substantial Revenue. Well, that's not exactly true. Comm Week, Jan 20 1997, "Content Will Ultimately Win the Game for ISPs" By Wayne Rash": I mentioned PSINet and Uunet last week, explaining that both companies were dumping their retail consumers and concentrating on business customers. As a result of this change in strategy, both companies are announcing profits for the last quarter of 1996, which is nearly unheard of in the Internet industry. >Revenue will cause Wall Street to favor your stock >even if you don't show real profits. When money are going to upgrades you can always get profit by slowing down sales. Like, by increasing prices or by dumping less profitable customers. That causes loss of market share, unless competitors are doing the slow-down act as well. Which they do (except maybe MCI). It's kind of pocker game with no limit... >The only company in the Internet >business that shows real profit is Cisco, and they are laughing all the way >to the bank. Of course vendors always have higher profit margin than >service companies. Well, most of Cisco's profit is coming from small boxes. _The_ laughing company is Intel... --vadim

1 0

RE: peering charges?
by Eric D. Madison 29 Jan '97

29 Jan '97

Well, that's not exactly true, UUNet nor PSI show "profits", they show substantial Revenue. Revenue will cause Wall Street to favor your stock even if you don't show real profits. The only company in the Internet business that shows real profit is Cisco, and they are laughing all the way to the bank. Of course vendors always have higher profit margin than service companies. Here is a financial statement for PSI: Their revenue is up tremendously but their bucket has a big hole in the bottom as does most other providers. Capital expenditure in this business is just so high. PSINet Inc. Corporate Performance 1995 1994 1993 1992 1991 ...................................................................... Revenues ($Mil) 38.7 15.2 8.7 6.4 4.7 Net Profit ($Mil) -53.2 -5.3 -1.9 0.4 0.2 EPS -1.78 -0.26 n/m n/m n/m Net Profit Margin (%) -137.3 -35.1 -22.1 6.7 5.0 UUNet info wasn't available because of the Merger.. their ticker doesn't show up anymore. I know they did not show profits because just before their merger I looked at all of the big players stock to see if there was anymone actually making money, and none has positive profits or margins. At 02:56 AM 1/29/97 -0800, you wrote: >>From Danny Stroud <dannystroud(a)msn.com>: > >>Don't harangue me about it. Ask Wall Street, look at the numbers for the ISPs, >>read the paper regarding the latest AOL flap. This is not my sole opinion. I >>will not engage in a philosophical debate on this issue. Sorry, but the >>market is the judge. des > >He. UUNET and PSI show profits even after lots of money spent >on expansion. Zillions of small ISPs which "resell" UUNET's and PSI's >services are profitable (or they'd close the shop long time ago, because >they generally have no cash reserves, and no VC funding). Sprint >and MCI do not show profits on Internet side, but their leased line >business is _very_ healthy because Internet sells those lines. > >If anything, telcos provide equivalemt of 64Kbps digital service to anyone >for the same $20/mo and are not particularly bankrupt. > >One has to be completely insane to call the industry with nearly decade of more >than 100% annual growth "not economically viable". One has to read way too much >creative economists (of statist camp) to believe in anything like that. For once, >those economists missed the entire new industry. Now they pretend it doesn't >exist or offer moronic suggestions on how to "fix" it. How nice, but we've >already seen exactly that scenario with PCs. > >--vadim > >PS AOL "flap" is a typical example of blatant incompetence. > Their marketroids never heard of market pilots, or something, > to measure response before doing something really drastic. > > If AOL will go under as a result, they won't get a lot of sympathy. > >PPS What figures? I've certainly seen a lot of creative accounting, > when, for example, network build-up is counted as operational > expense (Sprint used to do that), whereas all sane people count it > as capital. > >PPPS Wall Street who? I guess not the people who valued UUNET at $2B, > right? > 4 >P S May i politely suggest reading some books on the early history of > telephony before writing nonsense? Fortunately, they had fewer > economists and lawyers back then (though, Bell had to defend his > patents against 576 suits). > >

1 0

Domain names for ISP infrastructure links
by Peter Lothberg 29 Jan '97

29 Jan '97

So how do we name internal links in a way that is usefull for all operators, and conforms to the A-Z 0-9 and "-" and can be proposed as an RFC? --Peter

15 18

Re: peering charges?
by Vadim Antonov 29 Jan '97

29 Jan '97

Eric D. Madison wrote: >Since some of the larger vendors (Cisco mostly) has introduced accounting >features into their software settlements could start any time. a) the accounting was there for years, so what b) a 100-byte packet travelled from provider A to provider B. Should A pay to B or vice versa? So far nobody gave any useful answer to that question. There are no settlements because traffic has little relevance to relative worth of connectivity from one provider to another. The large ISPs are generally interested in market share or peers, not in volume of mutual traffic. --vadim

12 20

Internet Statistics
by Neal Castagnoli 29 Jan '97

29 Jan '97

Does anyone know a place where there are statistics kept on internet usage, such as Web vs. Mail vs. ftp? Thanks in advance, --Neal

2 1

comprehensive DNS scans from dao.andromedia.com
by Craig Nordin 29 Jan '97

29 Jan '97

Is anyone else seeing DNS ("ls") scan/requests from dao.andromedia.com ? I'm not sure I want to call it wrong, but it seems pretty cheesy to me to come over and try to check out all of my DNS systematically. I have them blocked, but I'm curious as to what they are up to. I asked once, and they said that they were analyzing the dial-up IP addresses that were hitting their web pages -- but they are slowly going over hundreds and hundreds of domain names under our control. There has to be something interesting in the story behind these scans. Even if it is only a weird elf :) -- Craig Nordin -- cnordin(a)vni.net Virtual Networks http://www.vni.net

3 3

Push's Tutorial
by Kent W. England 29 Jan '97

29 Jan '97

-----BEGIN PGP SIGNED MESSAGE----- Folks; Push's treatise on compelling content and content distribution was well thought and well presented. It may take some time to convince organizations that spend as much time at putting ink on paper and paper on driveways or shelves as they do in generating the content itself, that there is no place for content creators to participate in the distribution system. Of course, this presupposes the development of a real Web data architecture, as HTTP 1.1 evidences a direction. And it assumes that ISPs will follow through and create adequate business models to allow compelling content creators to buy distribution services easily and naturally. Push shows he, for one, has been thinking about how to do this. As an example, let us suppose that the NFL offers the ISP community a deal on carrying SuperBowl XXXII content. The SuperBowl is a good example (perhaps one of the best) of the web flash crowd phenomenon. But with the right upfront work whereby the NFL offers its terms and conditions for replication (for a fee) or caching (as a performance improvement), ISPs can sign on to carry the program. The NFL might offer ISPs a percentage of transactions if the ISPs carry the SuperBowl shopping cart. The NFL might offer the ISP a piece of the advertising space in exchange for carrying non-transaction content. The NFL might also offer terms and conditions, as for example, web hit stats for any ISP that wishes to cache certain SuperBowl content without remuneration. It's not too hard to envision that ISPs might eventually make money on content aggregation and delivery services and that many compelling content companies will be glad not to have to worry about the distribution channels. - - --Kent -----BEGIN PGP SIGNATURE----- Version: 4.5 iQCVAgUBMu6JBxvOn/xsDN5pAQGjtgP/XVEqRm9HYDc7tkq7rTCtagCQWNioO3SC rOGzZqIsH2K3370R2SZZa81U2ukyqRZQhY6H1Y51J4cj9jnbfWtDFKsEFvICInjn MF/xDrfU6bNNkLummaEOMM3jVzXBbO4hCPfxfeRl6zpeCfv5s7b1uXcvse4u+o68 W+TPNOn51U8= =GTrs -----END PGP SIGNATURE----- ~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~~~ Kent W. England Six Sigma Networks 1655 Landquist Drive, Suite 100 Voice/Fax: 619.632.8400 Encinitas, CA 92024 kwe(a)6SigmaNets.com Experienced Internet Consulting ~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~~~

2 1

None
by maxeyb＠mail.drsystems.com 29 Jan '97

29 Jan '97

1 0