- NANOG - lists.nanog.org

Re: 132.0.0.0/10 not in the databases
by Steven M. Bellovin 28 Nov '01

28 Nov '01

In message <5.1.0.14.2.20011128081413.00aa29f0@localhost>, Philip Smith writes: > >My theory is that DISO-UNRRA were originally allocated 132.1.0.0/16 through >132.15.0.0/16 in the classful world - these are all in the ARIN DB under >various military guises. When CIDR came along, it seems that someone must >have decided that because 132.0.0.0/16 was now available and part of a >bigger block, it could be added to the announcement, etc...? > >There are a total of four like this: > >Network Origin AS Description >132.0.0.0/10 568 DISO-UNRRA >135.0.0.0/13 10455 Lucent Technologies >137.0.0.0/13 568 DISO-UNRRA >158.0.0.0/13 568 DISO-UNRRA Umm -- how does Lucent fit into that? Last I checked, it wasn't part of DoD. Back in the mists of time, AT&T was allocated what we would now call 135.0.0.0/8. We allocated addresses according to what seemed like a rational scheme at the time, this being pre-CIDR. But a wandering neutron struck our CEO, inducing a fission event that produced (among other particles) AT&T and Lucent. 135.0.0.0/8 was split between the two companies as a collection of /16's, on the reasonably rational grounds of "whoever is using the block gets to keep it". This minimized disruption (or rather, avoided further disruption), at a time when there was plenty of other chaos involved in splitting companies, networks, buildings, and organizations. Unfortunately, it did not happen to correspond to CIDR principles, but as I said, the allocation to AT&T antedated CIDR and in no way anticipated what the CEO and the Board of Directors was going to do. >From 135.207.0.0/16, --Steve Bellovin, http://www.research.att.com/~smb Full text of "Firewalls" book now at http://www.wilyhacker.com

2 1

132.0.0.0/10 not in the databases
by Leo Bicknell 27 Nov '01

27 Nov '01

132.0.0.0/10 is floating around: route-views.oregon-ix.net>show ip bgp 132.0.0.0/10 BGP routing table entry for 132.0.0.0/10, version 12484641 Paths: (43 available, best #41) Advertised to non peer-group peers: 64.166.72.140 4513 13646 5696 568, (aggregated by 568 198.26.128.1) 195.66.224.82 from 195.66.224.82 (209.10.12.222) Origin IGP, metric 7602, localpref 100, valid, external, atomic-aggregate [snip, route-views has more than a page of entries for it] However, it doesn't apprar to be registered: % whois -h whois.arin.net 132.0.0.0 No match for "132.0.0.0". 568 regisitered to disa.mil, and a number of longer prefixes inside the /10 are registered to .mil addresses: % whois -h whois.arin.net 132.15.0.0 Kadena Air Base (NET-KADENA-NET) Bldg 792, Kadena AB JA APO, AP 96368-5149 US The question is, do they legitimately own the whole /10? If so, why isn't it marked in the databases? -- Leo Bicknell - bicknell(a)ufp.org - CCIE 3440 PGP keys at http://www.ufp.org/~bicknell/ Read TMBG List - tmbg-list-request(a)tmbg.org, www.tmbg.org

4 4

interliant / sagenetworks
by Dan Hollis 27 Nov '01

27 Nov '01

Does anyone have a confirmed working contact address for interliant / sagenetworks (AS 5697)? Ive gone through the usual channels and got no response from them. -Dan

1 0

RE: 132.0.0.0/10 not in the databases
by Borchers, Mark 27 Nov '01

27 Nov '01

This is all that shows up at http://www.nic.mil/: route: 132.0.0.0/10 descr: DISA, DSAC P. O. Box 1605 Columbus OH 43216-5002, USA origin: AS568 member-of: RS-COMM_US_FED mnt-by: MAINT-AS568 changed: nsfnet-admin(a)merit.edu 19940819 source: RADB > -----Original Message----- > From: Leo Bicknell [mailto:bicknell@ufp.org] > Sent: Tuesday, November 27, 2001 10:02 AM > To: nanog(a)merit.edu > Subject: 132.0.0.0/10 not in the databases > > > > > 132.0.0.0/10 is floating around: > > route-views.oregon-ix.net>show ip bgp 132.0.0.0/10 > BGP routing table entry for 132.0.0.0/10, version 12484641 > Paths: (43 available, best #41) > Advertised to non peer-group peers: > 64.166.72.140 > 4513 13646 5696 568, (aggregated by 568 198.26.128.1) > 195.66.224.82 from 195.66.224.82 (209.10.12.222) > Origin IGP, metric 7602, localpref 100, valid, > external, atomic-aggregate > > [snip, route-views has more than a page of entries for it] > > However, it doesn't apprar to be registered: > > % whois -h whois.arin.net 132.0.0.0 > No match for "132.0.0.0". > > 568 regisitered to disa.mil, and a number of longer prefixes > inside the /10 are registered to .mil addresses: > > % whois -h whois.arin.net 132.15.0.0 > Kadena Air Base (NET-KADENA-NET) > Bldg 792, Kadena AB JA > APO, AP 96368-5149 > US > > The question is, do they legitimately own the whole /10? If > so, why isn't it marked in the databases? > > -- > Leo Bicknell - bicknell(a)ufp.org - CCIE 3440 > PGP keys at http://www.ufp.org/~bicknell/ > Read TMBG List - tmbg-list-request(a)tmbg.org, www.tmbg.org >

1 0

Another national network has DSL problems
by Sean Donelan 27 Nov '01

27 Nov '01

These things seem to come in waves. Yet another large telco is having problems with their DSL service. Telstra Australia ADSL service went down about 9am local time (strange how these always happen at 9am, just as the dayshift starts working). Is this a trend, is there a common software fault or vendor these outages have in common? http://telstra.com/ServiceStatus/Default.asp

3 2

Re: how many roots must DNS have before it's considered broken (Re: ISP network design of non-authoritative caches)
by Simon Higgs 27 Nov '01

27 Nov '01

At 01:09 AM 11/22/01 -0500, Sean Donelan wrote: >I found an archive of NAMEDROPPERS going back to 1983. If you are >interested in the subject, or just want to know how we got here, its >an interesting read. > >Anonymous FTP at ftp://ops.ietf.org/lists/ in 1983, 1984, etc. > >If you don't know the past, you are doomed to repeat it. Even though >I disagree with some of them, I understand some of the decisions a >bit better. There are also some specific mailing lists such as newdom, where Draft Postel was hashed out. Due to various circumstances, the list moved a couple of times. There is an archive here (zip, gz, & mbx format): http://www.iahc.org/mail-lists.html Also there's the iahc-discuss mailing list, which unfortunately has been censored (between Jan 28, 1997 and May 14, 1997) because of the highly negative response to the IAHC's final report - i.e. the formation of eDNS on the IAHC mailing list: http://www.gtld-mou.org/gtld-discuss/mail-archive/ One of these days I'll get the shared-tld list archive online. Best Regards, Simon -- (A)bort, (R)etry, (T)ake down entire network?

1 1

dedicated NetFlow (V5&V8) probe for OC-3, OC-12 and OC-48
by Eddy Beier 27 Nov '01

27 Nov '01

Dear all, we've developed a NetFlow probe which processes about 1 Mpps (depends on the CPUs) unsampled traffic. We called it FlowBox. Details: - based on rackmount PC (1U/2U) - FreeBSD - PoS NICs with deep HW support for probing - AS list (RIPE format) loadable; so you don't need complete AS information on the router; we can also integrate a BGP reader if this is too static. - optional sending to multiple Collectors - optional passive (optical splitter) or active (lineloop/3R regeneration/media adaption) probing - optional V5 traffic reduction by a factor of about 1:10 by ignoring one (the always changing) TCP port; (still 100% NetFlow compliant) - tested by Cisco - tested with CAIDA's tools Problem: First we built the box, now we try to build a business plan for; but we failed up to now, 'cause there is no use for in europe (few number of high speed lines, accounting infrastructure not yet developed etc) Question: Is there anybody who thinks, that the box has been a good idea? Thanx in advance for any response (otherwise FlowBox will never touch the market). Yours Eddy Beier CEO bina GmbH Am Bruckweg 84 D-90499 Nuremberg ps: for the historically interested ones: we spinned off from the University of Erlangen with a complete Accounting system in a single box. with this box e.g. BMW account their european distributors (>800) and Allianz (europen insurance market leader) their agents (>8000) But it also had all problems of a complete accounting system in a single box (e.g. full file system, lack of analysis recources) and it's been proprietary. So we focused on the de facto accounting interface NetFlow and on probing.

1 0

Re: route-views.oregon-ix.net
by Marshall Eubanks 27 Nov '01

27 Nov '01

> >If somebody in this list is in charge of that, please contact me off-line. >I've been trying to contact dmm(a)antc.uoregon.edu but it have not received >any response. > >German Dear German; Try contacting David Meyer at <dmm(a)sprint.net>. > >~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ >* THE DILBERT FUTURE * >* Thriving on Stupidity in the 21st Century * >~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ >"Internet capacity will increase indefinitely >to keep up with the egos of the people using it" > > Marshall Eubanks tme(a)21rst-century.com

1 0

route-views.oregon-ix.net
by German Martinez 27 Nov '01

27 Nov '01

If somebody in this list is in charge of that, please contact me off-line. I've been trying to contact dmm(a)antc.uoregon.edu but it have not received any response. German ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ * THE DILBERT FUTURE * * Thriving on Stupidity in the 21st Century * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ "Internet capacity will increase indefinitely to keep up with the egos of the people using it"

1 0

Re: Next generation DSLAMs
by srihari varada 27 Nov '01

27 Nov '01

Read the posts by Geoff (inlcuding the provided article) and Miquel. I am trying to understand if ILECs could provision IP-VPN and MPLS-VPN service (not internet access) to, for instance, business subscribers with the ISPs, competitive and non-competitive, still providing internet access (and, perhaps Layer 2 or IP-VPN or MPLS-VPNs). Would this be prohibited by regulations? If so, under what context. Is it because, competitiveness would be hampered from the value added service provisioning point of view? Miquel had some references to it. Some elaborations would be useful. Secondly, is it reasonable to say that the notion of next generation DSLAMs a myth (until regulations change) from the perspective of provisioning IP-VPNs or MPLS-VPNs right in them. Regards, Srihari Varada

1 0