- Test - lists.nanog.org

Re: Inbound prefix filters
by Bradley Reynolds 12 Nov '97

12 Nov '97

On Tue, 11 Nov 1997, John A. Tamplin wrote: > Actually, I view it the other way. If someone is announcing routes for one > of our prefixes, connectivity is at least partially broken for that prefix, I think the whole point of filtering is that you will not send packets to that newly announced route. You can access-list it and if you suddenly see matching on some deny, then you can investigate at somewhere like nitrous and figure out who is announcing what. In general, I would not be willing to sacrifice the performance of the people who are paying me m0ney just to be able to quickly? ascertain who is causing the problem. > BTW, this has happened to us twice, and both times the offender was a > direct competitor in one of our local markets. Does anybody have any > feel for how often these "accidents" are not accidents? > Time to slap the kiddies for playing with daddy's router. brad reynolds ber(a)cwru.edu "Faith: not wanting to know what is true." -- Friedrich Nietzsche

1 0

IPv8 < IPv6
by Jim Fleming 12 Nov '97

12 Nov '97

Thanks for the responses to my IPv8 note. In case people missed the point, IPv8 addresses are smaller than IPv6. Here are the sizes. IPv4 - 32 bits IPv6 - 128 bits IPv8 - 43 bits (3+8+32) There is a natural routing hierarchy with IPv8 addressing....8 regions, 256 distribution centers in each region and full 32 bit Internets from there. IPv8 addresses can fit inside the IPv6 address fields. ==== Now...back to the point...NANOG people have the opportunity to help the Internet Community with Address Management and Governance. The specific protocol header is not nearly as critical as a coherent addressing plan that takes routing into consideration and quality of service. I suggest that people learn to manage 43 bit addresses before they try to deal with 128 bit addresses. It does little good to have huge addresses if people can not route packets from A to B because the Address Management is chaotic. NANOG people can help to bring order to the chaos that larger addresses can bring. Jim Fleming Unir Corporation IBC, Tortola, BVI

6 15

Hello? This one is wierd.
by Alex Rubenstein 12 Nov '97

12 Nov '97

Is this legal? traceroute to www.japan.net (204.68.24.180), 30 hops max, 40 byte packets 1 nac-nyc1-fa0-100Mb (207.99.0.201) 1.648 ms 1.063 ms 0.765 ms 2 e9-0.border1.nyc1.genuity.net (207.240.48.49) 1.569 ms 1.922 ms 4.960 ms 3 f6.core1.nyc1.genuity.net (207.240.1.81) 1.606 ms 1.737 ms 2.747 ms 4 core1.wdc1.genuity.net (207.240.0.25) 6.691 ms 8.987 ms 10.889 ms 5 h1.peer1.wdc1.genuity.net (207.240.1.246) 8.311 ms 10.539 ms 9.257 ms 6 maeeast.acsi.net (192.41.177.108) 8.230 ms 8.514 ms 8.878 ms 7 columb-sc-1-a12-0.acsi.net (206.222.97.1) 20.543 ms 22.417 ms 18.487 ms 8 greenv-sc-1-a12-0.acsi.net (206.222.97.5) 27.777 ms 27.418 ms28.025 ms 9 little-ar-1-a12-0-1.acsi.net (206.222.97.2) 60.061 ms 54.746 ms 55.946 ms 10 fortwo-tx-1-a12-0.acsi.net (206.222.97.6) 55.008 ms 61.942 ms 63.695 ms 11 904.Hssi4-0.GW1.HOU1.ALTER.NET (157.130.128.25) 60.870 ms 58.826 ms 61.469 ms 12 321.atm1-0-0.cr1.hou1.alter.net (137.39.21.109) 63.880 ms 57.682 ms 62.116 ms 13 112.ATM8-0-0.CR1.ATL1.Alter.Net (137.39.31.5) 201.734 ms * 72.052 ms 14 311.ATM6-0-0.SR1.ATL1.Alter.Net (137.39.21.73) 69.777 ms 71.955 ms 72.176 ms 15 h14-1.t104-0.Atlanta.t3.ans.net (140.223.65.18) 100.501 ms * 102.249 ms 16 h13-1.t64-0.Houston.t3.ans.net (140.223.65.13) 107.077 ms 99.397 ms 109.687 ms 17 h13-1.t80-1.St-Louis.t3.ans.net (140.223.25.14) 98.975 ms 99.187 ms 101.531 ms 18 h11-1.t24-0.Chicago.t3.ans.net (140.223.25.21) 115.362 ms 103.821 ms 93.777 ms 19 * h13-1.t96-0.Denver.t3.ans.net (140.223.25.18) 125.611 ms 125.056 ms 20 f1-0.c96-10.Denver.t3.ans.net (140.222.96.197) 123.094 ms 130.940 ms 124.054 ms 21 enss3336.t3.ans.net (207.25.224.18) 127.490 ms 123.043 ms * 22 * * * 23 * * *

5 4

IP Forwarding
by Tom Hubina 12 Nov '97

12 Nov '97

Hi all... I've been lurking here for a while, and realize that I really don't know much about the innards of networking, but I do have a question I was wondering if you could answer. If I'm in the wrong place, flame me, kick me, do whatever :) We have customers in Europe who need to download large (10 Meg+) files from us on an almost daily basis. They are all on ISDN or other low bandwidth dial-up so it takes them quite some time, if they are even succesful at all. At first glance a mirror site doesn't appeal to us since we have very frequent updates and need to provide security with frequent new user adds and changes. For now we've been living with the slow speeds and dropped transfers, but here's what I was wondering: Could a domain name be set up in Europe to point to a machine in Europe on a global service providers net (PSINet for instance). All this machine does is forward packets to a machine at our location on the same network (PSINet). The whole point of this would be to avoid overcrowded peer points where a lot of our current packetloss appears to be coming from. Does this idea have any merit, or would I simply be exchanging one problem for another? Would it even work at all? Is there a better solution that would increase our connection quality to Europe while keeping the server at our location? Thanks for your time. ---------------------------- Tom Hubina tomh(a)gibbsnc.com Gibbs and Associates 805-523-0004 ----------------------------

1 0

Re: Geographic v. topological address allocation
by Sean Donelan 11 Nov '97

11 Nov '97

Either geographic or topological address assignment will work. What works 'better' (I won't say 'best') depends on so-called externalities. In today's Internet the sender's provider keeps all the money. The sending provider wants the traffic off the sender's network as quickly as possible and onto the destination's provider network. In this case the goal is to minimize usage of the sending provider's network, so you want address assignment based on the destination provider, so you can find the nearest destination provider gateway from the source. On the other hand, lets assume a different economic model. If instead the sending provider had to pay the destination provider based on the mileage (e.g. V&H coordinates) between where the sending provider gives the traffic to the destination's provider and the final destination, the sender would want to carry the traffic as close to the destination as possible, assuming that its less expensive for the sender to use its facilities than to pay to use someone else's facilities, and hand the traffic to the destination's provider at the last possible gateway. In this case, the goal is to minimize mileage charges on the other provider's network, so you want the address assignment based on geography, so you can find the closest sending provider gateway to the final destination. The clunker in most models is traffic flows in both directions, and money also exists at both ends. Neither model is very good at global optimization of traffic or the money. Even having a single payer and single provider, such as was used in the NSFNET, resulted in sub-optimal use in some cases due to yet other externalities. What's interesting is as a provider's network reaches different sizes, different types of optimization become more desirable. And, yes, both styles of addressing can be used to make it difficult for competitors to enter the market.

2 1

better pay attention to holubs listing of folk at D o J
by Gordon Cook 11 Nov '97

11 Nov '97

herewith a few sentences from an AP story: Regulators are likely to examine the deal to see if such dominance would squeeze out rivals and drive up prices. But several analysts said they expect the deal to clear that hurdle, because there are roughly 4,000 other Internet access providers. ``They'll be a formidable competitor with a tremendous amount of market clout,'' said Rebecca Wetzel, a consultant with TeleChoice Inc., based in Verona, N.J. ``But there are plenty of other choices out there.'' ---------------- by saying there is nothing to worry about because there are 4,000 other internet access providers, the analysts betray their abysmal ignorance!!! I have a very unpleasant feeling that Sidgemore would like to be king of the internet. And that he likely doesn't know the internet well enough to realize that it rejects would be kings. If I were an ISP, I'd pay plenty of attention to David Holubs post of relevant people at DoJ to who to express your concerns. ************************************************************************ The COOK Report on Internet For subsc. pricing & more than 431 Greenway Ave, Ewing, NJ 08618 USA ten megabytes of free material (609) 882-2572 (phone & fax) visit http://cookreport.com/ Internet: cook(a)cookreport.com New Special Report: Internet Governance at the Crossroads ($175) http://cookreport.com/inetgov.shtml ************************************************************************

5 4

Current prefix length filtering
by Forrest W. Christian 11 Nov '97

11 Nov '97

I just recieved a new allocation from the internic. I had requested an /18 (based upon my previous allocation of an /18). The internic unfortunately has decided that an /19 is the largest block they can justify allocating to me. I don't want to get into an internic war here, although I am not that happy about the size of the allocation, as I feel that I've done everything right in conserving address space instead of doing what some smaller providers do and allocate everyone a /24, which would have consumed my /18 in 3 months instead of a year or so. The real question I have for nanog is who is filtering on 209/8 for prefixes longer than /18? I'm quite concerned about connectivity to the customers I'm going to have to put in this /19. Have all the major providers started filtering on /19's or am I going to have major problems with this block? Thanks, forrestc(a)imach.com

2 1

WorldCom - Schedule Maintenance (fwd)
by Eric D. Madison 11 Nov '97

11 Nov '97

Hmm.. So they finally started sending out scheduled outages for the Mae's.. Eric _______________________________________________________ Eric D. Madison - Senior Network Engineer - ACSI - Advanced Data Services - ATM/IP Backbone Group 24 Hour NMC/NOC (800)291-7889 Email: noc(a)acsi.net ---------- Forwarded message ---------- Date: Tue, 11 Nov 1997 13:18:07 -0600 From: Valerie Thomas <valerie.thomas(a)wcom.com> To: noc(a)above.net, madison(a)acsi.net, admin(a)agis.net, nc(a)ai.net, trouble(a)ans.net, reschly(a)arl.armt.mil, hostmaster(a)attmail.com, ttbrewster(a)attmail.com, mahmed(a)bactel.net, opc(a)bbnplanet.com, sysmail(a)cableinet.net, hostmaster(a)cableol.net, noc(a)cais.com, hostmaster(a)clr.com, hostmaster(a)conxion.net, noc(a)conxion.net, noc(a)cwi.net, noc(a)cyberconnection.com, bill_stritzinger(a)dataplace.net, dns(a)digex.net, noc(a)dra.net, admin(a)dx.net, abuse(a)erols.com, hosrmaster(a)es.net, noc(a)eu.net, simon(a)euro.net, support(a)exodus.net, kbacon(a)fibernet.net, ad-dormreg(a)genuity.net, noc(a)geo.net, hostmaster(a)getnet.com, jemmette(a)goodnet.com, tc(a)grid.net, bmanning(a)isi.edu, noc(a)primenet.com, estout(a)usgs.gov, help(a)uu.net, jun(a)wolfox.gip.net Subject: WorldCom - Schedule Maintenance ** Routed Mail. Please reply when completed ** Notification of Scheduled Maintenance - MAE EAST WorldCom Global Network Services has scheduled maintenance in Washington, DC on 11/13/97 at 23:00 et. until 24:00 et. We will swap out the linecards in slots 5 9 11 of giga6. This will reduce the huntgroup member density of each switch by a factor of one for the duration of approx 2min which includes the hard swap and spanning tree reconfig on a per card basis. All customers Mae wide will experience a slight degrade in service as there will be a temporary reduction in bandwidth. Some packet loss may be experienced. A maintenance ticket has been opened with our 24 hour Customer Service Center. The Ticket# is 138507. You may refer to these tickets for any questions or concerns related to this scheduled maintenance activity. The WorldCom/GNS Network Management Center will be monitoring all activity in conjunction with our local WorldCom Data Communications Consultant who will be on-site. The WorldCom staff will be readily available to ensure immediate attention is given to customer network service as required. If you should have any questions regarding this activity, please feel free to contact our Customer Service Center @ 1-800-MFS-7170. Customer Service

1 0

Summary: UUNET..
by jamie＠intuition.iagnet.net 11 Nov '97

11 Nov '97

I must commend UUNET for its response to my posting.. I haven't even received my post back to me that I sent this morning and already I have had three UUNET engineers (two from the US, one from Canada) _call_ me. The issue is: There's congestion between Tor-Cleveland. (Someone admitted it) :-) The resolution is: There are scheduled upgrades to OC3c the week of 11/17 from UU-CA to UU @ TCO. They're also adding more DS3s at other points. Thanks, all. -- jamie g.k. rishaw dal/efnet:gavroche __ IAGnet/CIC/netILLINOIS Backoffice DID:216.902.5455 FAX:216.623.3566 \/ 800.637.4IAGx5455 Chance favors the prepared mind.

1 0

RE: MCI Accepts Worldcom's Bid?
by Jim Browning 11 Nov '97

11 Nov '97

Say bye-bye to the concept of peering, huh? -- Jim ---------- From: Kevin Houle [SMTP:kevin@netins.net] Sent: Monday, November 10, 1997 6:43 AM Heard on the radio this morning that MCI accepted Worldcom's latest bid of ~$36 billion. MCI's website confirms. http://www.mci.com/aboutus/company/news/wcom/index.shtml Kevin

14 15