- NANOG - lists.nanog.org

Re: more-specifics in class B space?
by Sean Donelan 23 Mar '00

23 Mar '00

On Wed, 22 March 2000, Joe Provo - Network Architect wrote: > It also appears that there are parties who received address delegations > within classical B space and are parceling chunks out, without any > delegation or registration. Without that administrativia, and no clear > customer-vendor relationship in an AS path, how does this behavior look > any different to address hijacking? This was the primary reason why I didn't listen to more specifics than normal ARIN CIDR delegations. I was willing to punch a hole for anyone who could come up with a good story, but almost all the announcements were mistakes or hijacks. Someone unknown person would announce a /17 out of a university's address block, and black hole half of the university. This is a severe, but silent, problem in today's Internet. These hijackings have resulted in worse DoS attacks than the Yahoo! floods in February, but because so few NOCs know what to look for, most people just think its the flacky Internet. Yes, I'm aware of AS 702. It still doesn't make it right. My question is why can't you announce an aggregate matching the allocated address block? If you have no common AS path, no common administrative authority, no common provider, why are you using the same network block? In every case I've come across so far, the correct answer is either get seperate blocks or announce an aggregate at your border or your upstream's border. This is the usual answer in the case of a state university with several campuses. They may have more specifics within their network and their upstreams, but the upstream announces a common aggregate for the entire state network. Or if they are truely independent institutions they should get independent address blocks from one of the address registries on an appropriate boundary. This is a different issue than getting the registries to allocate small CIDR blocks.

1 0

Re: Sprint routing issues
by trall＠almaden.ibm.com 23 Mar '00

23 Mar '00

On 03/22/2000 at 23:34:54 EST, jlewis(a)lewis.org wrote: > Is it just me or is Sprint having issues tonight? I noticed some routing > problems getting to C&W hosted sites via Sprint. When I tried to go to > www.sprint.net, I couldn't resolve it. According to F, sprint.net's NS's > are: > > NS1-AUTH.SPRINTLINK.net. 2D IN A 206.228.179.10 > NS2-AUTH.SPRINTLINK.net. 2D IN A 144.228.254.10 > NS3-AUTH.SPRINTLINK.net. 2D IN A 144.228.255.10 > > These are all currently unreachable from our site (directly connected to > Sprint and UUNet) and multiple offsite locations on other backbones. I can ping all of them, but none answer DNS queries. Makes it tough to access their customers. Tony Rall

4 4

IETF-47
by Alexei Roudnev 23 Mar '00

23 Mar '00

May be, someone from here will be in Adelaida at the IETF-47 meeting?

1 0

Sprint routing issues
by jlewis＠lewis.org 23 Mar '00

23 Mar '00

Is it just me or is Sprint having issues tonight? I noticed some routing problems getting to C&W hosted sites via Sprint. When I tried to go to www.sprint.net, I couldn't resolve it. According to F, sprint.net's NS's are: NS1-AUTH.SPRINTLINK.net. 2D IN A 206.228.179.10 NS2-AUTH.SPRINTLINK.net. 2D IN A 144.228.254.10 NS3-AUTH.SPRINTLINK.net. 2D IN A 144.228.255.10 These are all currently unreachable from our site (directly connected to Sprint and UUNet) and multiple offsite locations on other backbones. If I traceroute from atlantic.net (goes out via Sprint) to my own site (www.lewis.org) hosted off a C&W T1, I get 2 hops into C&W, and then nothing. If I trace from www.lewis.org to atlantic.net, I get 2 hops into Sprint, and then nothing. Before anyone asks...yes, I did call Sprint and run this all by them...and they opened a ticket...but the fact that seemed more interested in doing intrusive testing on our T3 than looking at their own network didn't instill much confidence. ---------------------------------------------------------------------- Jon Lewis *jlewis(a)lewis.org*| Spammers will be winnuked or System Administrator | nestea'd...whatever it takes Atlantic Net | to get the job done. _________http://www.lewis.org/~jlewis/pgp for PGP public key__________

2 1

Re: more-specifics in class B space?
by Jim Mercer 23 Mar '00

23 Mar '00

On Wed, Mar 22, 2000 at 05:13:45PM -0700, Marc Slemko wrote: > Any examples of such providers? So far, shaw.ca is the main one I found. ah, that is exactly the same one i had issues with. they also said something about sprint, but the sprint looking glass was seeing the sub-/16 subnets that were being advertised. one wonders if shaw.ca is the only one adhereing to "ARIN guidelines". if so, i'd like some ammunition to go whack 'em in the forehead with. -- [ Jim Mercer jim(a)reptiles.org +1 416 506-0654 ] [ Reptilian Research -- Longer Life through Colder Blood ] [ Don't be fooled by cheap Finnish imitations; BSD is the One True Code. ]

1 0

viz interface to cflowd
by k claffy 22 Mar '00

22 Mar '00

there was mention on this list last week of cflowd http://www.caida.org/Tools/Cflowd/ not having much in way of viz API here's another knight that's done some cool graphics interface to cflowd with tobi's RRDtool fwiw (hey check out the napster binning: http://net.doit.wisc.edu/data/Napster/ ) k ---------------------------------------------------------------------- From: Dave Plonka (plonka(a)doit.wisc.edu) Date: Tue Mar 21 2000 - 13:59:40 PST To: cflowd(a)caida.org Cc: dwm(a)caida.org, oetiker(a)ee.ethz.ch Subject: Re: FlowScan-1.002 released X-Mailer: Mutt 0.95i On Wed, Mar 22, 2000 at 02:04:21PM -0800, k claffy wrote: I'm pleased to announce the release of FlowScan-1.002. FlowScan is a tool to monitor and graph flow information from Cisco routers in "near" real-time. I've pushed myself to release it now since I've just recently added passive measurement of Napster traffic. A sample of what FlowScan can do is at: http://wwwstats.net.wisc.edu Availability FlowScan is licensed under the GNU General Public License, and is available to you at: http://net.doit.wisc.edu/~plonka/FlowScan/ Mailing Lists There are two mailing lists having to do with FlowScan: * flowscan a general mailing list for FlowScan users. * flowscan-announce a low-volume, restricted post mailing list to keep FlowScan users informed of news regarding FlowScan. Announcements will be "cross-posted" to both lists, so there's no need to join both. These lists are hosted by the Division of Information Technology's Network Engineering Technology group at the University of Wisconsin - Madison. To subscribe to either of them, send email to: majordomo(a)net.doit.wisc.edu containing either: subscribe flowscan *or*: subscribe flowscan-announce You should receive an automatic response that will request that you verify your request to become a member of the list, to which you must reply with the authentication information there-in. Then, in response to your reply, you should receive a welcome message. If you have any questions about the administrative policies of this list's manager, please contact: owner-flowscan(a)net.doit.wisc.edu *or*: owner-flowscan-announce(a)net.doit.wisc.edu Napster If you're interested in Napster specifically, note the graph labeled "Well Known Services Mb/s" here: http://wwwstats.net.wisc.edu or, directly, here: http://wwwstats.net.wisc.edu/CampusIO/services_Mbps.gif I have written a general document about the Napster portion of the measurement at: http://net.doit.wisc.edu/data/Napster/ Thanks FlowScan-1.001, which was essentially the beta version, has been tested quietly by a number of volunteers since September of last year. I'd like to thank the beta testers for their efforts and feedback. Also, thanks to Daniel McRobb, Tobi Oetiker, and CAIDA for providing the main tools upon which FlowScan is built, namely "Cflowd" and "RRDTOOL". -------------------------------------------------------------

1 0

Re: Cisco's CNR vs ISC Bind/DHCP
by Ron Buchalski 22 Mar '00

22 Mar '00

Fletcher writes: >From: Fletcher E Kittredge <fkittred(a)sss1.gwi.net> >To: nanog(a)merit.edu >CC: gwi-network(a)sss1.gwi.net >Subject: Cisco's CNR vs ISC Bind/DHCP >Date: Thu, 16 Mar 2000 17:02:37 -0500 > >We are working here with the hypothesis that Cisco's CNR product is a >lamer and inferior in quality to the Internet Software Consortium's >open source DNS and DHCP software. Is this a generally accepted >hypothesis, or are we off in the weeds (again) by ourselves on this >one? > >thanks! >fletcher Well, Fletcher, considering that you probably don't know anything about the product, do you really feel it necessary to publicly bash it? On what information to you base your bold assertion? Would you appreciate it if people had similar assumptions about: http://www.gwi.net/ ??...... Go to: http://www.cisco.com/warp/public/cc/cisco/mkt/enm/cap/cnr/index.shtml and download a free (limited-time) copy of CNR to check it out. -rb ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com

6 7

Lucent AccessPoint 450
by Oleg Shpak 21 Mar '00

21 Mar '00

Does anybody have an experience with Lucent AccessPoint 450 ? How reliable this equipment ? Are you pleased to use it ? I have looked for Cisco7204, but AccessPoint looks fine too, especially its price. I need to connect company office to two poviders via bursted T3 links. I need to do load balancing between these links in both direction. It must do bgp, possible ospf, access-lists, maybe traffic shaping. Thanks, Oleg Shpak.

1 1

non-op
by James Paul 21 Mar '00

21 Mar '00

I apologize if this is off topoic I am researching the deployment of a free personal web-hosting service. Normally each user's html documents reside in a subdirectory of the document tree where the subdirectory is named for the user. /htdocs/ / \ / \ user1 userN Thus a users URL is www.site.com/userN. We currently have about 5000 users on the service. This is spread across three different single directory filesystems. As the number of users grows I imagine that this will become unworkable due to the nature of the file systems. One solution is to put groups of users in several subdirectories but this could make the user URLs messy. www.site.com/sub1/sub2/userN Can someone please comment on methods of keeping short clean user URLs while making the storage of many small web-pages more efficient and scaleable? What kind of backends are currently being used to serve up plain old static content with inline stuff? Also, how are directories being indexed for retrieval? How have are the larger installations currently working this? Thanks in advance, James Paul

2 1

FCC's NRIC to monitor ISP outages
by Robert Cannon 20 Mar '00

20 Mar '00

NRIC V had its first meeting today. NRIC is a Federal Advisory Counsil providing information to the FCC on telecommunications (ie telephone) network reliability. Information on NRIC can be found at www.nric.org The Chair of NRIC is Jim Crowe from Level3. The primary focus of this NRIC will be to continue the network reliability work of the group, including packet switched networks. NRIC will be broken into the following subgroups (1) Y2K - finishing the work (2) Network Reliability and (3) Interoperability. NRIC IV recommended that a one year voluntary trial period of network outage reporting be initiated for a wide variety of communications service providers, including Internet Service Providers. The final report of NRIC IV's recommendations in this area can be found at http://www.atis.org/atis/nricgr3.htm The relevant language from the Executive Summary is below. This Report was placed on the web for public comment in February for public comment. Those comments were received and incorporated. The recommendation has been accepted by NRIC. Thus, presumably, NRIC will move forward to initiate a one year voluntary trial period of ISPs report outages to NCC/NCS, NRIC, and ultimately to the FCC. At the completion of the trial, an analysis will be done of the usefulness of the reporting. NRIC indicates that the issue of ISP involvement in the voluntary reporting was supported by Ira Richter of IOPS www.iops.org The presentations from todays NRIC meeting should soon be on the NRIC website www.nric.org Exec Summary: "A voluntary trial is recommended with participation by service providers of CMRS, satellite, cable, data networking and INTERNET SERVICE PROVIDERS (ISPs) to alert National Communications Systems/National Coordinating Center for Telecommunications (NCS/NCC) of outages that are likely to have a significant public impact. "* Industry associations should provide an informational notice to their membership to inform them of a voluntary outage reporting trial and encourage their participation. "* Analysis of the data from this voluntary trial should be done by a neutral party. This analysis should be similar to scope to the analysis conducted on wireline carrier segments. "* At the completion of a voluntary trial period (minimum 1-year) an evaluation of the effectiveness of the data for usefulness to participants and the FCC should be understaken. "Industry communication of NRIC IV 'Best Practices' to CMRS, Satellite, cable, ISP, data networking service providers is recommended." -B www.cybertelecom.org cannon(a)world.oberlin.edu Views Dey Are My Own ______________________________________________ FREE Personalized Email at Mail.com Sign up at http://www.mail.com/?sr=signup

1 0