February 2001 - Test - lists.nanog.org

Re: Using unallocated address space
by Sean Donelan 15 Feb '01

15 Feb '01

On Thu, 15 February 2001, Omachonu Ogali wrote: > So now we need someone to hold the hand of 17,000 ASes because of 1 > or 2 mishaps. Rethink this please. I doubt the DFZ needs a government. > If you hate it, filter them, if not ignore it. It is considerably more than 1 or 2 mishaps. However, unless it happens to www.whitehouse.gov, it doesn't get much attention. On any day, if you looked at the route table, you would find several examples.

1 0

[CoralReef 3.4.1 released to public]
by k claffy 15 Feb '01

15 Feb '01

----- Forwarded message from Ken Keys <coral-info(a)caida.org> ----- Date: Thu, 15 Feb 2001 11:55:53 -0800 From: Ken Keys <coral-info(a)caida.org> To: coral-dev(a)caida.org Cc: elfcore(a)caida.org, "Joerg B. Micheel" <joerg(a)cs.waikato.ac.nz> Subject: CoralReef 3.4.1 released to public CoralReef 3.4.1 is now available. Both public and member versions can be found at http://www.caida.org/tools/measurement/coralreef/. CoralReef is a comprehensive software package from CAIDA for passive monitoring of ATM, POS, and other network interfaces and reading "crl" and pcap tracefiles. It includes FreeBSD drivers for Apptel POINT (OC12 and OC3 ATM) and FORE FATM (OC3 ATM) cards, support for WAND DAG (OC3 and OC12, POS and ATM) cards on linux, programming APIs for C and perl, and software applications for capture, analysis, and reporting of ATM, IP, and TCP/UDP traffic. Direct questions to coral-info(a)caida.org. Major differences between the last public release and this one include: * Support for live pcap interfaces * Support for DAG capture cards and file formats * Support for POS, CHDLC, PPP (over POS, ATM, or Ethernet), Bridged ethernet over PPP * Tcpdump style packet filtering in all CoralReef packet applications * t2_report++ can graph the top N applications (members only) * Many other enhancements and fixes to applications and C and perl APIs For a complete list of changes, see http://www.caida.org/tools/measurement/coralreef/doc/doc/CHANGELOG. -- Ken Keys CoralReef: http://www.caida.org/tools/measurement/coralreef/ ----- End forwarded message -----

1 0

Using unallocated address space
by Sean Donelan 15 Feb '01

15 Feb '01

On Mon, 12 February 2001, John Fraizer wrote: > Any time a network is caught announcing non-allocated address space, the > registry should bill them accordingly. If they refuse to pay, the > registry should yank their ASN. That would be strong encouragement to do > the right thing. Other than making it difficult for people to figure out WHOIS using that ASN, "yanking" an ASN's registration has little practical effect. You can use an un-allocated ASN almost as easily as using an un-allocated address block.

6 8

cablemodems and voip
by Miguel Mata-Cardona 15 Feb '01

15 Feb '01

<color><param>0100,0100,0100</param><FontFamily><param>System</param>Hi, I've been following this list quite a while, I'm a little off topic but in the need of serious help. You can mail me direct all your help and I'll post a summary at the end. thanks in advance. ----------------------------------------------- problem: We are a cablemodem provider and have disabled the peer to peer options in our CMTSs to avoid problems with NBT and directed broadcast to our customers. The network has been split in many nodes around the city and each node is as follows: to other nodes | | ------------- | router | | 10.2.15.1 | ------------- | | --------------- | cable modem | | termination | | system | --------------- /\ / \ / \ / \ me other 10.2.15.2 10.2.15.3 As you might guess, under this configuration "me" and "other" won't be able to ping to each other. We have done some tests with linux setting "route -net 10.2.15.0 netmask 255.255.255.0 gw 10.2.15.1" and it works but it does not on win9x. We are looking for this solution because we are deploying a VoIP to some customers on the same node. The above solution, if possible under win9x, would bring a problem because we are modifying the customer machine instead of our network but is feasible solution anyway. Is there a way on doing this within our nodes? if not, is there a way to modify or tweak the win9x routing table?<FontFamily><param>Arial</param> <nofill> -------------------------- Miguel Mata-Cardona CCI El Salvador mmata(a)sv.cciglobal.net

2 1

Juniper M series/Cisco GSR comparison
by Martin Cooper 15 Feb '01

15 Feb '01

Matthew Correll writes: > I'm looking for objective sources for a comparitive analysis > between Juniper's M10 and/or M20 routers and Cisco's GSR 12000 > series. [...] > I'm not looking for opinions or vendor bias - only documented > testing done by reliable (and hopefully objective) sources. Cisco appears to have commisioned an independent evaluation of its GSR 12016 series by "The Tolly Group" for what it's worth: <URL:http://www.cisco.com/warp/public/cc/pd/rt/12000/prodlit/tolly_in.pdf> M.

1 0

RE: BGP deployment and peering questions
by Richard Welty 15 Feb '01

15 Feb '01

Ryan O'Connell wrote: > Unless Cisco have changed something, a 3640 can only take > 64Mb of memory max. gee, then how come we have more than 64M in ours, and they work? it's the 3620 that can only take 64M, the 3640 can take up to 128M. richard

1 0

NANOG
by Martin Hannigan 15 Feb '01

15 Feb '01

Well, I hate to do it, but there's no way in hell I can go. Turns out I picked up a HUGE HUGE client thanks to Pam Frailey and the XMAS party. Speaking of that, who wants to start the pool whether I can have an OC-12 up and running and pushing 400mbps of juice in 6 weeks? Joe Provo can't join he may have inside data! -M Regards, -- Martin Hannigan hannigan(a)fugawi.net Fugawi Networks Engineering Boston, MA http://www.fugawi.net Ph: 617.742.2693 Fax: 617.742.2300

1 1

Re: BGP deployment and peering questions
by Ryan O'Connell 15 Feb '01

15 Feb '01

On Thu, Feb 15, 2001 at 10:49:28AM +0100, Blaz Zupan wrote: > > Unless Cisco have changed something, a 3640 can only take 64Mb of memory max. > > You'd need a 3660 for 128Mb. > > You got that backwards. A 3640 can take 128Mb of memory just fine, a 3620 is > limited to 64Mb. Yes, you're quite right of course. I should know that, I have 3640s with 96Mb in. (Must have more Caffiene before NANOGing...) And, of course, it's 256Mb for a 3660. -- Ryan O'Connell - <ryan(a)complicity.co.uk> - http://www.complicity.co.uk I'm not losing my mind, no I'm not changing my lines, I'm just learning new things with the passage of time

1 0

Re: Using unallocated address space
by Roy 15 Feb '01

15 Feb '01

Nice list but.... 1) only covers /8 2) Doesn't seem to be maintained (look at 32/8). Roy Engehausen Jasper Wallace wrote: > On Tue, 13 Feb 2001, Roy wrote: > > > > > > > What would be nice is that there was an offical list of private, unallocated, or > > reserved ranges kept somewhere that we could use and generate filters from. > > http://www.isi.edu/in-notes/iana/assignments/ipv4-address-space > > -- > Internet Vision Internet Consultancy Tel: 020 7589 4500 > 60 Albert Court & Web development Fax: 020 7589 4522 > Prince Consort Road vision(a)ivision.co.uk > London SW7 2BE http://www.ivision.co.uk/

2 1

Re: Using unallocated address space - for DoS?
by Sean Donelan 14 Feb '01

14 Feb '01

On Wed, 14 February 2001, "Miguel A.L. Paraz" wrote: > On Wed, Feb 14, 2001 at 09:21:32AM -0500, Steven M. Bellovin wrote: > > You don't have to break into the "right" router; you just have to start > > announcing the networks in a way that your peers don't -- can't -- > > detect is improper. > > I did not mean that the network operator was malicious. I meant, in the same > way that vulnerable servers are broken into and used for DoS, can routers > be broken into and do DoS via blackholes? I think it is hard unless you > know the right combination of vulnerable router (sniffable LAN?) and > unprotected upstream or peer. Once again, you don't need to break into someone else's routers. This isn't an exploit list, so I'm not going to post a cookbook on how to do it. But due to the limited protection in some networks, it remains a relatively simple attack or accident waiting to happen. Why break into a bank, when you can push a button on the side of the building and money comes out. Unlike a cash machine, where you need to know at least a PIN in addition to some working bank account number, you can wipe out almost any IP address you don't like with essentially no authentication.

1 0