January 2007 - Test - lists.nanog.org

Opentransit route-server
by German Martinez 16 Jan '07

16 Jan '07

Hello, After few days experiencing problems, we have upgraded the hardware for our route-server. telnet://route-server.opentransit.net ipv6: 2001:688:0:3:4::5 Comments are welcome Thanks German Martinez

1 0

Comment spammers chewing blogger bandwidth like crazy
by Thomas Leavitt 16 Jan '07

16 Jan '07

A friend of mine operates a blog at seeingtheforest.com, and he pays for traffic over a (fairly minimal) cap. He posted this comment recently: http://www.seeingtheforest.com/archives/2007/01/eating_bandwidt.htm Eating Bandwidth Last month something ate up a tremendous amount of bandwidth at Seeing the Forest, costing me a lot of money. So now I regularly check bandwidth use. Why has 209.160.72.10, HopOne in DC, been eating a HUGE amount of bandwidth? Gigabytes! What are they doing? (I banned them.) Why has 220.226.63.254, an IP in India, been eating a tremendous amount of bandwidth? What are they doing? Why has 195.225.177.46, an IP in Ukraine, been eating a tremendous amount of bandwidth? What are they doing? Why has 62.194.1.235 AND 83.170.82.35 AND 89.136.115.220 AND 62.163.39.183 AND 212.241.204.145, all from the /same company/ in Amsterdam, been eating a TREMENDOUS amount of bandwidth? What are they doing? Why is 206.225.90.30 and 69.64.74.56 and Abacus America Inc.eating a TREMENDOUS amount of my bandwidth, *** One of the comments said: Yeah, I've seen a huge bump in my blog's traffic, I haven't figured out what they're doing, but it ate like 4Gb of bandwidth last month. Now that you mention it, I checked last month's stats and yep, there's 209.160.72.10 producing 62% of my blog traffic. I did a little checking around the web and they're an obvious spam host. Banned. *** They also chew up a lot of CPU (comment filter code). At few times, myself, I've had to simply take code offline that was getting hit too heavily... seems like the IPs (and their ilk) listed above are good prospects for a "bad behavior" blacklist, at a level below that of "collaborative spam filter" (which doesn't prevent traffic or CPU cycles from being consumed). Given the volume of traffic mentioned, this must be a real problem for some hosts and networks... although, on the other hand, if their marginal use rates are high enough, they might actually be making money off this. Regards, Thomas Leavitt -- Thomas Leavitt - thomas(a)thomasleavitt.org - 831-295-3917 (cell) *** Independent Systems and Network Consultant, Santa Cruz, CA ***

15 20

oracle net ops contact
by Dmitry Kiselev 16 Jan '07

16 Jan '07

Hello! Could anybody from AS794 (Oracle) NOC contact me out of the list? I see routing issue inside AS794. Thanks. -- Dmitry Kiselev

1 0

Re: AFP article on Taiwan cable repair effort
by Robert Bonomi 16 Jan '07

16 Jan '07

> Date: Tue, 16 Jan 2007 15:04:17 +0100 > From: "D.H. van der Woude" <dirkvanderwoude(a)gmail.com> > To: NANOG(a)merit.edu > Subject: Re: AFP article on Taiwan cable repair effort > > Thanks for the link, which brought me to the page > http://www1.alcatel-lucent.com/submarine/vessels/index.htm > where their vessels are decribed. > > Their specs bring up something (at to least to me) of interest. > The popular story is always "in the 90-ties so much submarine > (trans-Atlantic & -Pacific fiber was laid that it will take decades > to fill it up. And the bust of 2000 seemed proof enough. > > However, from the specs of Alcatel's 5 vessels one learns > that 4 of them are effectively from 2002 (!). > > So did Alcatel some anti-cyclic investing - or did they know > more than others? 1) Any idea how long it takes to manufacture one of those vessels? Any idea how long in advance of start of manufacture that the 'firm order' had to be placed? 2) Were they additions to Alcatel's fleet, or _replacements_? 3) With all that additional cable 'in place', the amount of maintenance required goes up proportionately.

1 0

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 14 Jan '07

14 Jan '07

Any insight will be appreciated. regards, /virendra

1 0

Prefix list formats: advice needed
by Bill Woodcock 14 Jan '07

14 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 14 Jan '07

14 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 13 Jan '07

13 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