February 2001 - Test - lists.nanog.org

www.networksolutions.com off the air today?
by Jeff Frost 14 Feb '01

14 Feb '01

Can anyone verify that Network Solutions is off the air today or that its just a problem on my side of the wire? I haven't been able to reach the website all day and can't ping it either. Traceroute follows: mail:/home/jeff% traceroute www.networksolutions.com traceroute to www.networksolutions.com (216.168.224.69), 30 hops max, 40 byte packets 1 216.32.111.193 (216.32.111.193) 1 ms 1 ms 1 ms 2 216.35.106.114 (216.35.106.114) 1 ms 1 ms 1 ms 3 dcr03-g2-0.sntc03.exodus.net (216.33.152.97) 1 ms 1 ms 1 ms 4 bbr02-g4-0.sntc03.exodus.net (216.33.153.66) 1 ms 1 ms 1 ms 5 mae-west-ames-p1-0-0.exodus.net (216.32.173.54) 2 ms 2 ms 2 ms 6 * * Of course they might block ICMP.

4 4

OSU finally plans to fix their open relays!
by John Fraizer 14 Feb '01

14 Feb '01

After years of having tons of SPAM go through their servers, knowing about it and refusing to close their open relays, OSU has decided that perhaps they want to be able to send email and not have it bounce with a MAPS or ORBS message. They flat out admit that they had their servers set as open relays INTENTIONALLY in the following release: http://www.oit.ohio-state.edu/newsroom/openrelay.html Not that I don't appreciate this greatly but, OSU: It's about %^#%^&# time! --- John Fraizer EnterZone, Inc

3 2

[bgp-stats] Routing Table Report (fwd)
by Travis Pugh 14 Feb '01

14 Feb '01

It must be "leak private ASN" day. According to apnic's routing analysis, abovenet is leaking private ASNs like a sieve. route-views.oregon-ix.net shows a large leak from Telstra, and I've got Qwest leaking private ASNs into my local tables. * i65.132.120.0/24 209.227.130.34 50 100 0 209 65209 i * i 209.227.130.34 50 100 0 209 65209 i *>i 209.227.134.34 50 100 0 209 65209 i * i65.133.67.0/24 209.227.130.34 50 100 0 209 65209 i 65535 PRIVATE 62.4.68.96/27 6461 AboveNet Communicati 65053 PRIVATE 64.124.5.128/26 6461 AboveNet Communicati 65519 PRIVATE 64.124.60.0/24 6461 AboveNet Communicati 65519 PRIVATE 64.124.61.0/24 6461 AboveNet Communicati 65056 PRIVATE 64.124.72.128/26 6461 AboveNet Communicati 65526 PRIVATE 64.124.79.64/27 6461 AboveNet Communicati 65041 PRIVATE 64.124.98.0/24 6461 AboveNet Communicati 65041 PRIVATE 64.124.99.0/24 6461 AboveNet Communicati 65041 PRIVATE 64.124.104.0/24 6461 AboveNet Communicati 65041 PRIVATE 64.124.105.0/24 6461 AboveNet Communicati 65044 PRIVATE 64.124.160.192/26 6461 AboveNet Communicati 65046 PRIVATE 64.124.161.0/26 6461 AboveNet Communicati 65049 PRIVATE 64.124.170.64/27 6461 AboveNet Communicati Did someone hose up their "remove-private-as" function in software? -travis

2 1

Re: /24's run amuck?
by Danny McPherson 14 Feb '01

14 Feb '01

> Umm... because if you filter the /24s you will remove 58K of 95K, > or some 61%. I'm *sure* that having a routing table 40% of the original > size will help the next time you have a BGP flap. Of course, some attentive customers MAY notice reachability issues with those 58K prefixes, but no big deal, at least the routing table is smaller. Heck, might as well go ahead and filter /23 and longer, that'll loss another ~10K, or better yet... :-) -danny

4 3

Re: Juniper M series/Cisco GSR comparison
by smd＠clock.org 14 Feb '01

14 Feb '01

objective on this issue? hahahahahahahahahaha! no no, you have to figure out how to sort out the truth from conflicting hyped-to-the-max, name-calling, tribalistic rants. juuniper sux, they smell like pine trees, scott's crazy, pradeep's paranoid... Sean. (wondering if he still has friends there after this, or if they actually agree :) )

1 0

Juniper M series/Cisco GSR comparison
by Correll, Matthew 14 Feb '01

14 Feb '01

I'm sure I've missed something on this in the past, but I'm looking for objective sources for a comparitive analysis between Juniper's M10 and/or M20 routers and Cisco's GSR 12000 series. Specifically, I'm interested in the pros/cons of deploying each of these boxes at an AS boundary in a large service provider network. Some areas of particular interest to me include: 1. Fault Tolerance - Ability to recover from hardware component failures 2. BGP Robustness - Ability to process high levels of BGP updates without affecting forwarding performance; stability of routing code; interoperability between vendor implementations 3. CoS features - performance during congestion; results of different queueing and congestion control approaches 4. Interface performance - performance of individual line cards (OC-12 and OC-48 PoS) 5. Total system performance - aggregate switching capacity results under load I'm not looking for opinions or vendor bias - only documented testing done by reliable (and hopefully objective) sources. Thanks Matt Correll

1 0

Test on Cisco 720xVXR NSE-1
by Michael Kim 14 Feb '01

14 Feb '01

Hi! I have done some performance test on Cisco 720xVXR NSE-1 and the result I got in my test do not match with the performance test done by Cisco. Is anybody done test on 720xVXR NSE-1? /michael --

2 1

Abovenet BGP upgrade problems
by jfielding＠ixl.com 14 Feb '01

14 Feb '01

On Sunday night, Abovenet upgraded their BGP sessions and ever since then I've been receiving an incomplete BGP table. They sent a nice message telling me of the upgrade(see below) but it appears that they wiped out 10,000 networks from their BGP advertisements. The networks included www.microsoft.com, nitrous.digex.net, and a /15 client network. It's been causing problems for my network and I've had a ticket open with Abovenet for over 24 hrs. They added the networks that I complained about to their advertisement, but I don't know how to get a diffs of the 10,000 networks that I'm missing. The output from my BGP summary is... 93736 network entries and 187404 paths using 15463992 bytes of memory The output from nitrous.digex.net shows... 106558 network entries and 528694 paths using 28942878 bytes of memory Are any other Abovenet customers experiencing the same problems? Thanks, Jay Fielding iXL, Inc. Network Engineer -----Original Message----- From: Christine Jordan [mailto:cjordan@above.net] Sent: Friday, February 09, 2001 4:15 PM To: tech-contact(a)ns.above.net Subject: Network Wide Upgrades Starting Monday, February 12, 2001 Dear Valued Customer, Starting this coming weekend and continuing for several weeks, we will be implementing network-wide changes that will be of tremendous value to all of our customers moving forward. The changes will have minimal impact of our customers. Starting on Monday, February 12, 2001, we will be updating customer BGP sessions. We will start in Atlanta (ATL1), Virginia (IAD2, IAD4), New York (LGA1), Chicago (ORD1), Seattle (SEA1), Portland (PDX1) and San Jose (SJC2). Impact to our customers should be minimal on this. On Tuesday, February 13, 2001, we will continue with San Jose (SJC1) and Virginia (IAD1). As we progress, we will keep you informed of all work being done. If you have any questions regarding this technical contact, please contact our 24x7 Call Center by email at service(a)above.net or call 1-877-ISX-PERT / 1-408-350-6451. Thank you. AboveNet Communications Metromedia Fiber Network

3 2

Re: Using unallocated address space - for DoS?
by Steven M. Bellovin 14 Feb '01

14 Feb '01

In message <20010214215512.A19817(a)mail.q-linux.com>, "Miguel A.L. Paraz" writes : > >On Tue, Feb 13, 2001 at 09:47:15PM -0800, Sean Donelan wrote: >> Unfortunately this is not a unique occurance. Cable&Wireless, Sprint, >> AT&T and UUNET have all had portions of their service knocked off the >> Internet for various periods of time due to bogus announcements. Until >> other ISPs fix their policies, I can knock your network off most of the >> Internet, and there is nothing you can do to prevent it. > > >I hope this is a remote possibility, but what are the chances of someone >malicious breaking into the "right" router and blackholing the worst possible >networks? If this is done, how long till it can be remedied? 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. --Steve Bellovin, http://www.research.att.com/~smb

2 1

RE: Using unallocated address space
by Sean Donelan 14 Feb '01

14 Feb '01

On Tue, 13 February 2001, "Brett L. Hawn" wrote: > Here I go being silly again, but how about people take responsability for > their own networks and filter properly at their borders? All this talk of > how to enforce things is pretty meaningless when you have countless members > of NANOG itself half-assing their own networks and complaining about other > people's. Because this is only half the answer. I always filtered my announcements, was careful to register all the address blocks I used, and was very responsible for my own network. It had no effect on someone else hijacking one of my addresses and announcing it through a large ISP's route tables. I was effectively cut off the network not because of anything I did, or could control. Worse there was little I could do to fix it. I had to wait three days for the large ISP (with whom I had no direct relationship) engineer's to decide it was worth their effort to stop the source of the false announcement. Unfortunately this is not a unique occurance. Cable&Wireless, Sprint, AT&T and UUNET have all had portions of their service knocked off the Internet for various periods of time due to bogus announcements. Until other ISPs fix their policies, I can knock your network off most of the Internet, and there is nothing you can do to prevent it.

3 2