- NANOG - lists.nanog.org

Another article on the fiber cut - a geopolitical slant
by Suresh Ramasubramanian 15 Jan '07

15 Jan '07

Pointing out some of the chokepoints in international undersea fiber routes (the straits of malacca, luzon straits etc) and discussing the geopolitical implications of these http://www.pinr.com/ has the article but as it has a "no redistribution without written permission" disclaimer .. go there anyway. Worth a read. -- Suresh Ramasubramanian (ops.lists(a)gmail.com)

1 0

recommendation for a co-lo in bombay (india).
by virendra rode 15 Jan '07

15 Jan '07

Any insight will be appreciated. regards, /virendra

1 0

Prefix list formats: advice needed
by Bill Woodcock 15 Jan '07

15 Jan '07

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Howdy. For a tool we're writing, we need to be able to accept lists of prefixes people might want to BGP advertise. We'd like to be "liberal in what we accept" and recognize lists formatted in as many ways as people think they might reasonably have. There are some obvious ones: - Delimited lists of CIDR-format addresses. This would include addresses in "slash" notation, with some punctuation or whitespace between them. That would cover a lot of hand-typed and tool- generated lists, as well as Juniper-format prefix lists. - One-address-per-line, CIDR-format, with other non-address stuff in the line. This would include Cisco ip prefix-list format, most IRR whois, and some RIR whois. - One-address-per-line, with separate address mask in normal order (ones values on the left, zeros values on the right), with other non- address stuff in the line, some of which might look like more addresses. This would include Cisco ip route statements. - One-address-per-line, with separate address mask in reverse order (zeros values on the left, ones values on the right), with other non-address stuff in the line, some of which might look like more addresses. This would include Cisco ip access-list statements. - Address-ranges, characterized by a starting address and an ending address, perhaps with a hyphen between them. This would include many older RIR whois entries. Can anyone think of any other formats people are likely to have lists of prefixes in? Thanks much, -Bill -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.3 (Darwin) iD8DBQFFqqrfGvQy4xTRsBERAt/7AKClUXCT3N5NJBGVReildfMO/xQjMACcDDDH 1EEYeKq71UiayuppDXHu3Yo= =uQ03 -----END PGP SIGNATURE-----

2 1

NANOG 39 BOF: Pushing the FIB limits, perspectives on pressures confronting modern routers.
by Joel Jaeggli 15 Jan '07

15 Jan '07

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Greetings, We've been working on putting together a BOF for NANOG 39 titled: Pushing the FIB limits, perspectives on pressures confronting modern routers My intent with this BOF was to gather the interested parties to look at when current router/switch architectures run out of TCAM or SRAM on the line-cards (depending on flavor) to hold the fib, and potentially what's being done about it. The scalability of BGP has been presentation fodder for a couple NANOG meetings now, ultimately though running a router hinges on having the data you need in the fib. Comparatively buying more ram for, or scaling your rp is a problem, but not an immediately intractable one. There are a number of forces putting pressure on the FIB, IPV4 de-aggregation and multi-homing, traffic engineering, intra-provider vpn engineering, etc. Buying new hardware is obviously one approach, but some operators have stated to me that they are pursuing alternatives such as aggressive filtering. If you are going to be at NANOG 39 and you have an interest in the subject or perspective you'd like to offer please attend the BOF. If you'd like some time on the agenda, mail me and we'll add you. Thanks Joel Jaeggli -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.4 (MingW32) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFFqtOL8AA1q7Z/VrIRAqBpAJ0d/AonXwzmQni22zwtnKjOWuEWigCeK1Fg QngJ9c0pXxaFu377paCpxjA= =I29s -----END PGP SIGNATURE-----

1 0

Re: ISP Security BOF @NANOG 39
by MARLON BORBA 14 Jan '07

14 Jan '07

If I could attend that BOF (I am in Brazil :-/)... Either way I'd like to suggest a subject:- "Whois: Latest Developments At ICANN And How They Affect Incident Response" Abraços, Marlon Borba, CISSP Técnico Judiciário - Segurança da Informação TRF 3ª Região (11) 3012-1683 -- Segurança é um processo, não um produto. -- Bruce Schneier -- >>> Randy Bush <randy(a)psg.com> 01/13/07 4:19 PM >>> i wanna hear/see Defending the NANOG net. i.e. how the local net is geared for security. i am especially interested as we will be needing to do a lot more for the AfNOG net this year, and could use the tech transfer. randy

1 0

Re: Network end users to pull down 2 gigabytes a day, continuously?
by odlyzko＠dtc.umn.edu 13 Jan '07

13 Jan '07

This is the case of bundling, discussed in the paper I referenced in the previous message, http://www.dtc.umn.edu/~odlyzko/doc/history.communications1b.pdf It is impossible, at least without detailed studies, to tell what the effect of selling individual channels would have. Bundling can have benefits for both consumers and producers (and that is what the cable industry in the US claims applies to their case, although all we can conclude for sure from their claims is that they believe it has benefits to them). Here is a simple example of bundling (something that has been known in standard economics literature for about 30 years, although in practice this has been done for thousands of years in various markets): >From what Marshall wrote, it appears that the 2 channels that he and his family care about are worth at least $40 in total to him, and everything else is useless. Suppose (and this may or may not be true) he and his family value each of these channels, call them A and B, at $30/month and $20/month, respectively, so in principle the cable network could even raise their bundles' prices to a total of $50 without losing him as a subscriber. Now suppose that the universe of users consists just of Marshall and Mikael, except that Mikael and his family are interested in 3 channels, the two channels A and B that Marshall cares about, and channel C, and suppose the willingness to pay for them is $10, $5, and $25, respectively. If the cable company has to price the channels separately (and let's exclude the ability to price discriminate, namely charge different prices to Marshall and Mikael, something that is generally excluded by local franchise agreements), what will they do? They will surely ask for $30 for channel A, $20 for channel B, and $25 for channel C, and will get $50 from Marshall and $25 from Mikael, for a total of $75. On the other hand, if all they offer is a bundle of all 3 channels for $40/month, both Marshall and Mikael will pay $40 each for a total of $80/month. And note that both Marshall and Mikael will be getting the bundle for no more (less in Marshall's case) than their valuations of individual components. If $75/month is not enough to pay the content providers and maintain the network at a profit, the lack of bundling may even lead to death of the network. Andrew P.S. And don't forget that having channels is already a form of bundling, as are newspapers, ... > On Sat Jan 13, Marshall Eubanks wrote: On Jan 13, 2007, at 7:36 AM, Mikael Abrahamsson wrote: > > On Sat, 13 Jan 2007, Marshall Eubanks wrote: > >> A technical issue that I have to deal with is that you get a 30 >> minute show (actually 24 minutes of content) as 30 minutes, _with >> the ads slots included_. To show it without ads, you actually have >> to take the show into a video editor and remove the ad slots, >> which costs video editor time, which is expensive. > > Well, in this case you'd hopefully get the show directly from > whoever is producing it without ads in the first place, basically > the same content you might see if you buy the show on DVD. > I do get it from the producer; that is what they produce. (And the video editor time referred to is people time, not machine time, which is trivial.) >> In the USA at least, the cable companies make you pay for >> "bundles" to get channels you want. I have to pay for 3 bundles to >> get 2 channels we actually want to watch. (One of these bundle is >> apparently only sold if you are already getting another, which we >> don't actually care about.) So, it actually costs us $ 40 + / >> month to get the two channels we want (plus a bunch we don't.) So, >> it occurs to me that there is a business selling solo channels on >> the Internet, as is, with the ads, for order $ 5 - $ 10 per >> subscriber per month, which should leave a substantial profit >> after the payments to the networks and bandwidth costs. > > There is zero problem for the cable companies to immediately > compete with you by offering the same thing, as soon as there is > competition. Since their channel is the most established, my guess > is that you would have a hard time succeeding where they already > have a footprint and established customers. > Yes, and that has the potential of immediately reducing their income by a factor of 2 or more. I suspect that they would compete at first by putting pressure on the channel aggregators not to sell to such businesses. (note : this is NOT a business I am pursuing at present.) What I do conclude from this is that the oncoming wave of IPTV and Internet Television is going to be very disruptive. > Where you could do well with your proposal, is where there is no > cable TV available at all. Regards > > -- > Mikael Abrahamsson email: swmike(a)swm.pp.se

1 0

Re: Comment spammers chewing blogger bandwidth like crazy
by Gregory Hicks 13 Jan '07

13 Jan '07

> Date: Sat, 13 Jan 2007 18:58:02 +0000 (GMT) > From: "Chris L. Morrow" <christopher.morrow(a)verizonbusiness.com> > Subject: Re: Comment spammers chewing blogger bandwidth like crazy > To: Thomas Leavitt <thomas(a)thomasleavitt.org> > Cc: nanog <nanog(a)merit.edu> > > > > On Sat, 13 Jan 2007, Thomas Leavitt wrote: > > > Why has 195.225.177.46, an IP in Ukraine, been eating a tremendous > > amount of bandwidth? What are they doing? > > this isn't in the ukraine, it's in NYC behind ISPrime. Phil is fairly > hhelpful, you might ask them to 'figure out what the heck is going on' > with that ip :) > > -Chris > (unless the ukraine got a whole lot closer to IAD than I thought: > 64 bytes from 195.225.177.46: icmp_seq=1 ttl=55 time=13.1 ms > 64 bytes from 195.225.177.46: icmp_seq=2 ttl=55 time=24.5 ms Um-m-m-m... % Information related to '195.225.176.0 - 195.225.179.255' inetnum: 195.225.176.0 - 195.225.179.255 netname: NETCATHOST descr: NetcatHosting country: UA admin-c: VS1142-RIPE tech-c: VS1142-RIPE status: ASSIGNED PI mnt-by: RIPE-NCC-HM-PI-MNT mnt-lower: RIPE-NCC-HM-PI-MNT mnt-by: NETCATHOST-MNT mnt-routes: NETCATHOST-MNT notify: vs(a)netcathost.com changed: hostmaster(a)ripe.net 20040304 source: RIPE remarks: *************************************** remarks: * Abuse contacts: abuse(a)netcathost.com * remarks: *************************************** person: Vsevolod Stetsinsky address: 01110, Ukraine, Kiev, 20Á, Solomenskaya street. room 206. phone: +38 050 6226676 e-mail: vs(a)netcathost.com nic-hdl: VS1142-RIPE changed: vs(a)netcathost.com 20040303 source: RIPE > ) --------------------------------------------------------------------- Gregory Hicks | Principal Systems Engineer Cadence Design Systems | Direct: 408.576.3609 555 River Oaks Pkwy M/S 9B1 San Jose, CA 95134 I am perfectly capable of learning from my mistakes. I will surely learn a great deal today. "A democracy is a sheep and two wolves deciding on what to have for lunch. Freedom is a well armed sheep contesting the results of the decision." - Benjamin Franklin "The best we can hope for concerning the people at large is that they be properly armed." --Alexander Hamilton

4 3

Re: Network end users to pull down 2 gigabytes a day, continuously?
by odlyzko＠dtc.umn.edu 13 Jan '07

13 Jan '07

Extensive evidence of the phenomenon Mike describes (inexpensive, frequently used things moving towards flat rate, expensive and rare ones towards sophisticated schemes a la "Saturday night stop-over fares") is presented in my paper "nternet pricing and the history of communications," Computer Networks 36 (2001), pp. 493-517, available at http://www.dtc.umn.edu/~odlyzko/doc/history.communications1b.pdf It also explains some of the mechanisms behind this tendency, drawn both from conventional economics (bundling, etc.) and behavioral economics (willingness to pay more for flat rates). This tendency can indeed reverse in cases of extreme asymmetry of usage. But one has to be careful there. Heavy users are often the most valuable. (In today's environment they are often the ones who provide the P2P material that attracts other uses to the network. And yes, there is a problem there, in that you don't need such heavy users to be on YOUR network for them to be an attraction in signing up new subscribers.) Andrew > On Sat Jan 13, Mike Leber wrote: On Sat, 13 Jan 2007, Sean Donelan wrote: > On Fri, 12 Jan 2007, Stephen Sprunk wrote: > > There is no technical challenge here; what the pirates are already doing > > works pretty well, and with a little UI work it'd even be ready for the mass > > market. The challenges are figuring out how to pay for the pipes needed to > > deliver all these bits at consumer rates, and how to collect revenue from all > > the viewers to fairly compensate the producers -- both business problems, > > though for different folks. > > Will the North American market change from using speed to volume for > pricing Internet connections? Web hosting and other markets around the > world already use GB/transferred packages instead of the port speed. The North American market started with charging per GB transfered and went away from it because the drop in cost per Mbps for both circuits and transit made costs low enough so that providers could statistically multiplex their user base and offer "unlimited" service (unlimited for marketing departments is being able to offer something to 99 percent of your customer base, which explains all residential service clauses that state unlimited doesn't really mean unlimited). You can see this repeatedly for all sorts of products as costs have come down in the long view. For example, consumer Internet dialup, long distance calling plans, local phone service plans, some aspects of cell phone service, it might be happening with online storage right now (i.e. google gmail/gfs and the browser plugins that let you store files in your gmail account). What might or might not be trending is a digression, the "unlimited" service is a marketing condition that seems to occur when 99 percent of your customer base uses less than the cost equal to the benefit of offering "unlimited" service. Mike. +----------------- H U R R I C A N E - E L E C T R I C -----------------+ | Mike Leber Direct Internet Connections Voice 510 580 4100 | | Hurricane Electric Web Hosting Colocation Fax 510 580 4151 | | mleber(a)he.net http://www.he.net | +-----------------------------------------------------------------------+

1 0

Re: Comment spammers chewing blogger bandwidth like crazy
by Alexander Harrowell 13 Jan '07

13 Jan '07

I was asked to join late in 2005. On 1/13/07, Eric Brunner-Williams in Portland Maine < brunner(a)abenaki.wabanaki.net> wrote: > > Do you operate fistfulofeuros? That's a good blog/community. > > I operate wampum and koufax, and draftgore2008, and we do see > persistant commerical ad inserts, and the occasional event for > which no commercial motive is self-evident. > > Eric >

1 0

Re: Comment spammers chewing blogger bandwidth like crazy
by Gregory Hicks 13 Jan '07

13 Jan '07

> Date: Sat, 13 Jan 2007 20:14:23 +0000 (GMT) > From: "Chris L. Morrow" <christopher.morrow(a)verizonbusiness.com> > > On Sat, 13 Jan 2007, Gregory Hicks wrote: > [...] > > > > > > this isn't in the ukraine, it's in NYC behind ISPrime. Phil is fairly > > > hhelpful, you might ask them to 'figure out what the heck is going on' > > > with that ip :) > > > > > > -Chris > > > (unless the ukraine got a whole lot closer to IAD than I thought: > > > 64 bytes from 195.225.177.46: icmp_seq=1 ttl=55 time=13.1 ms > > > 64 bytes from 195.225.177.46: icmp_seq=2 ttl=55 time=24.5 ms [...] > yes, but 'whois info' is not often 'correct' especially in this case, > traceroute to it, unless ISPrime has some time-space vortex that ip > (that one of the /22) is actually in NYC. speed-o-light don't often lie... Yup!. Should have looked further. Chris is right. (Learn from the master! - I *should* have known better...) Behind at least two levels of ISPrime... Thanks! --------------------------------------------------------------------- Gregory Hicks | Principal Systems Engineer Cadence Design Systems | Direct: 408.576.3609 555 River Oaks Pkwy M/S 9B1 San Jose, CA 95134 I am perfectly capable of learning from my mistakes. I will surely learn a great deal today. "A democracy is a sheep and two wolves deciding on what to have for lunch. Freedom is a well armed sheep contesting the results of the decision." - Benjamin Franklin "The best we can hope for concerning the people at large is that they be properly armed." --Alexander Hamilton

1 0