- NANOG - lists.nanog.org

Re: Core router bakeoff?
by Randy Bush 07 May '98

07 May '98

> Wait.. a Bay BCN will push 1 Gb/s through the box, problem is to WHERE they will push those bits. yes, bay makes nice hardware. unfortunately, routers are really software. and real routers move a bunch of that software to asics. randy

1 0

Re: NANOG 13 Logistics Update
by Howard C. Berkowitz 07 May '98

07 May '98

DHCP - you bet. Interstingly enough, there was a recommendation at one point to have DHCP address assignment free, but charge $3 for a piece of paper with a static IP address. Then someone else suggested a competitive market model was needed, with multiple static-IP-address-on-paper vendors to allow the market to produce the most efficient paper static address producers. A fraction of the $3/address would then go into an allocation fund. This allocation fund (until congress gets involved) would be then be used to lower the cost of cookies at NANOG. Bill At 04:07 PM 5/7/98 -0400, Paul Ferguson wrote: >At 03:18 PM 5/7/98 -0400, William B. Norton wrote: > >> >>Connectivity at NANOG 13 >>------------------------ >>Bring your laptops and cables! We will have over 400 ports available plus >>wireless connectivity including 40 wired tables. If you need to configure >>your firewalls in advance... >> What lookup algorithm do the tables use? Cisco's new trie? If a hash, corned beef, roast beef, red flannel, or what? As to Jamie Rishaw's comment, >Please fill out the isp-ip-cookie-template.txt located at http://rs.arnc.net. >The American Registry for Nanog Cookies is a non-profit organization. > >from http://rs.arnc.net/feeschedule.html : >Your Cookie Fee will be based on the total allocation of cookies received >at the previous NANOG. Users receiving cookies for the first time will be >charged a fee based on the number of chocolate chips in your initial cookie. >The cookie allocation fee will include inverse addressing (toilet-paper) >service, updates and maintenance. Are we having a problem of competitive registries for NANOG cookies, magic cookies, and browser cookies?

1 0

Re: NANOG 13 Logistics Update
by William B. Norton 07 May '98

07 May '98

DHCP - you bet. Interstingly enough, there was a recommendation at one point to have DHCP address assignment free, but charge $3 for a piece of paper with a static IP address. Then someone else suggested a competitive market model was needed, with multiple static-IP-address-on-paper vendors to allow the market to produce the most efficient paper static address producers. A fraction of the $3/address would then go into an allocation fund. This allocation fund (until congress gets involved) would be then be used to lower the cost of cookies at NANOG. Bill At 04:07 PM 5/7/98 -0400, Paul Ferguson wrote: >At 03:18 PM 5/7/98 -0400, William B. Norton wrote: > >> >>Connectivity at NANOG 13 >>------------------------ >>Bring your laptops and cables! We will have over 400 ports available plus >>wireless connectivity including 40 wired tables. If you need to configure >>your firewalls in advance... >> >>Address space: 207.75.182.0-207.75.183.0 >> >>We will also have a small number of PCs running a flavor of unix in a >>terminal room down the hall. >> >>We will have IPv6 connectivity, IPv6 DNS, and a Squid web cache server on >>site. >> > >DHCP? > >- paul > > > -------------------------------------------------------------- William B. Norton <wbn(a)merit.edu> (734) 764-9430 (New Area Code)

5 4

NANOG 13 Logistics Update
by William B. Norton 07 May '98

07 May '98

Hi all - Bunch of logistics information about the upcoming NANOG in Dearborn... NANOG 13 Agenda --------------- As usual, we have a very full agenda. See: http://www.nanog.org/mtg-9806/topics.html for details. I'd like to especially point out that we are starting the tutorials early this time. We have a Sunday early afternoon tutorial "Introduction to RPSL" (Cengiz Alaettinoglu, ISI, Curtis Villamizar, ANS ). In addition, we will be going all the way up to 4PMish on Tuesday, so book your flights accordingly. NANOG Registration ------------------ We will be cutting off registration at 500 again. Please register at: http://www.nanog.org If you are using NorthWest airlines, note that the NW machinists contract negotiation difficulties have resulted in a work slowdown, causing occasional flight delays and cancellations. You may want to take this into account scheduling your flights - one strategy is to fly in on Saturday AM and return Tuesday AM. Hotel Room Block Extended until May 15th ---------------------------------------- The NANOG hotel block reservation has been extended until May 15th, after which time the hotel room cost will rise from the NANOG rate of $125/night to the normal rate of $165/night. Make your reservations directly with the Hyatt, mentioning the NANOG meeting: Hyatt Regency Dearborn Hotel Fairlane Town Center Dearborn MI 48126-2793 Telephone: 1 313 593 1234 Sales Fax: 1 313 593 3366 If you run into problems, contact Pam Ciesla (Merit) at 734-936-0172 or e-mail her at pam(a)merit.edu. Connectivity at NANOG 13 ------------------------ Bring your laptops and cables! We will have over 400 ports available plus wireless connectivity including 40 wired tables. If you need to configure your firewalls in advance... Address space: 207.75.182.0-207.75.183.0 We will also have a small number of PCs running a flavor of unix in a terminal room down the hall. We will have IPv6 connectivity, IPv6 DNS, and a Squid web cache server on site. For those who can not attend, the meeting will be cast over both the Real Broadcast Network (RBN) as well as to the MBone for those that use vic/vat. RBN has graciously agreed to store the meeting on disk for later on-demand replay as well. We will ask the speakers provide web pointers in advance and link them into the agenda for those attending remotely (the camera shots of the screens never work well). Looking for Monday Evening BOFs ------------------------------- We have 3 rooms and the main ballroom Monday evening to use, so we are looking for Birds-of-a-Feather session proposals from folks. If you want to suggest a BOF or lead one, please send me a paragraph or so describing the BOF and the audience. These can be groups of interest, coordination discussions, emerging technology discussions. If possible, list members of the community that share an interest in airing the topic. There will be a preference for BOFs which have not been done at NANOG in the past. The only stipulation for these BOFs is that these *must not* be commercial product/service promotions. Fair warning: The community will respond very harshly to blatant commercial pitches. T-Shirts Welcome ---------------- While we do walk a fine line on the commercial influences at NANOG, noone has ever complained about free T-shirts at NANOG! To this end, if your company would like to ship boxes of T-shirts to the meeting, there should be no shortage of takers! If you are interested, send me e-mail (wbn(a)merit.edu) Beer-n-Gear Monday Evening -------------------------- Aptis, Bay and Cisco have graciously sponsored the beer-n-gear event Monday evening, immediately following the meeting. The kegs will be tapped around 5PM, dig into thousands of dollars worth of overpriced hotel food, and enjoy some lively technical discussions over some of the latest gear. (My experience is that these discussions get only livelier with the addition of beer ;-) ) Sponsorships ------------ The NANOG counts on the support of the community for everything from agenda topics to loaner equipment and event sponsorship. We would like to thank the following companies for their commitments to NANOG: +Bay Networks for providing the room connectivity. We will have dozens of hubs for over 400 wired ethernet ports for this meeting! +RBN for casting the meeting and providing storage and reply facilities +Cisco for providing the hardware and staff to cast the meeting on the MBone +University of Michigan Information technology Division for providing PCs for the terminal room +Aptis, Bay, and Cisco for participating in the Beer-n-Gear event +Cabletron for providing Wireless transceivers and a handful of loaner PCMCIA cards +MichNet for providing transit connectivity to the Hyatt Dearborn While the beer-n-gear event is currently full, we are still looking for companies to sponsor (or co-sponsor) breaks, breakfasts, or lunches for the meeting. If you would like more information, please send e-mail to me (wbn(a)merit.edu) with "NANOG 13 SPONSORSHIP" in the subject line or call (734-764-9430). Want to help set up equipment Sunday Morning? --------------------------------------------- If you are around Sunday morning and want something to do, help us set up routers and hubs, run wires, set up the terminal room, etc. Stop down to the Hubbard Ballroom and lend a hand - we have lots to set up in just a few morning hours, so we could really use the help! That's all I have. Cheers - Bill -------------------------------------------------------------- William B. Norton <wbn(a)merit.edu> (734) 764-9430 (New Area Code)

2 1

Latest Cisco IOS release with Version 5 flow export?
by dougd＠mail.airmail.net 07 May '98

07 May '98

Okay the cisco people haven't answered this in three days. *sigh* Does anyone know what the latest version of Enterprise IOS to support exporting Version 5 flow export data? i.e. the `ip flow-export version` command The one I am running is 11.1(18)CA and there is a bug in it which I need to upgrade out of. The 11.3 code they want us to go to only exports version 1 code. (or they've moved the command somewhere else) If you private email the answer to me I will summarize to the list.

2 1

Re: BGP & CIDR blocks
by Sean Donelan 07 May '98

07 May '98

> A bit of sanity..Prefix based filtering != not caring about your >backbone. Quite the reverse. As far as trusting bgp goes, most times you >can, sometimes you can't. Prefix filtering is for the times you >can't. Everybody remembers that those times happen, right? I remember, although it is getting to be a problem with each provider having a 'not invented here' syndrome. I give Sprint a list of our prefixes in one format, MCI a list of our prefixes in another format. I'm not a big fan of centralized storage, since whois.ra.net WHOIS server doesn't seem to be responding tonight. But it would be nice if I could generate the information once, and let all the providers use it.... Gosh, I wonder what that could be. > If your AS is in an appropriate (your upstream providers) autnum, >and the route object for the prefix is in the <pick the registry of your >choice> things work without a phone call. When they don't work, our noc >does a good job of helping to figure out what the problem is and getting >it fixed.. That exception is the process Jon is in. Exactly how this works hasn't been very clear to an outsider. I've had customers with their own AS number, who moved from one provider to another provider. I added their ASN to our AS-MACRO in the RADB and announced their AS and network prefixes. When their first provider stopped announcing the AS, the customer couldn't reach ANS. A call to the ANS NOC received a response in addition to updating the RADB we were supposed to send ANS a seperate message telling you the AS number was now being announced by our network (as indicated by the AS-MACRO). New network prefixes, and new AS numbers seem to get in Ok. But moving AS numbers still seems to be a special case. I can understand why, but it would be nice if there was some way you could let other ISPs know what the special cases are, so we don't get surprised. If this is not the case, I know I'm not alone in appreciating knowing when or not we have to send ANS a seperate note for AS number policy changes. On a related subject, is there any concern about cleaning obsolete objects out of the various routing registries? Information entropy seems out of control in most of these registries. It looks like they could use a good dose of authority control and weeding. If you aren't a library cataloger I suspect you've never heard of it. Or should we just declare the registry databases a failed experiment, and leave them as trash dumps for the archalogists to dig through. -- Sean Donelan, Data Research Associates, Inc, St. Louis, MO Affiliation given for identification not representation

3 2

Re: BGP & CIDR blocks
by dharvey＠dataplace.net 07 May '98

07 May '98

We just had the same problem with ANS. It took them almost two weeks to address the issue. They even told me that it was fixed several time, even though it had not been. Drake Harvey Manager, Internet Technologies DPNet.Net dharvey(a)dataplace.net Is it unreasonable of me to think it's awfully backwards of ANS.NET to require anyone getting a new CIDR block to physically call ANS's NOC and say "please listen to routes for this block from this AS"? First they told me "we don't trust BGP, so we use the various routing registries to build our routing tables." So I got our new block and our AS in radb.ra.net. A few days (and an ANS routing table update or two) go by, and still we have no connectivity to ANS. Then I get an email from them asking me to please call thier NOC to verify that they really should listen to our route. Isn't the point of BGP that you don't have to manually take care of routing updates? Are there other backbones I need to call to make sure they see our routes? ------------------------------------------------------------------ Jon Lewis <jlewis(a)fdt.net> | This space reserved for routing Network Administrator | update confirmations. Florida Digital Turnpike | ______http://inorganic5.fdt.net/~jlewis/pgp for PGP public key____

1 0

Re: Routing Policy and http://rs.arin.net/ip-allocation.html
by Michael Gibson 07 May '98

07 May '98

As part of Canadian Internet Community, we have been following the CIDR guidelines. We also do not route IPs from other CIDR blocks. Michael Gibson Sr. Network Admin - Netcom Canada -----Original Message----- From: Michael Dillon <michael(a)memra.com> To: nanog(a)merit.edu <nanog(a)merit.edu> Date: Wednesday, May 06, 1998 5:13 PM Subject: Re: Routing Policy and http://rs.arin.net/ip-allocation.html >On Wed, 6 May 1998, Mark Kent wrote: > >> That is the wrong question to ask. The real question is who has been >> handing out 206.116/16 like they were dealing cards? A quick look >> at the announcments and in whois.arin.net reveals a bunch of /24 >> spread out over Canada. > >This is because historically, addresses in Canada were handed out without >regard to CIDR even after everyone else was doing CIDR allocations. A >couple of years ago the Canadian IP registry was shut down and allocations >since that time have followed CIDR guidelines more closely. If you want to >see an excellent example of the kind of things that used to happen, have a >look at "whois -h whois.arin.net 199.166.227". In 1994 I applied for a >CIDR block of two Class C's, i.e. a /23 and this is what I got. Look >closely now... If you still can't see the problem try writing both /24's >in binary one above the other. > >> Historically, and currently, the Canadians crowd the top of >> the Tony Bates cidr list. Why is that? > >Part of it is because of the above allocation problem but another large >part of the problem is a severe lack of clue. Not enough Canadians go to >NANOG meetings to learn how to run a network properly, especially at the >Canadian backbone providers. Even when there were 10 provincial backbones >within the CA*NET consortium, the knowledge transfer just did not take >place. In some provinces like Ontario the network operators seemed more >interested in socialist politics than in running a network properly. > >Things are a bit better now in some regions and in some companies but >there is still a lot more knowledge sharing that could be done. > >And if anyone wants to flame me for this, please don't inflict it on the >NANOG list. Instead, come see me in person this weekend at the 12th annual >Canadian Internet conference http://www.net98.bc.ca and tell me what a >twit I am. I'll even buy you a beer. > >-- >Michael Dillon - Internet & ISP Consulting >http://www.memra.com - E-mail: michael(a)memra.com > > >

1 0

VIX Looking Glass (was: AS1853 announcing default)
by Christian Panigl, ACOnet/UniVie 07 May '98

07 May '98

>Date: Wed, 06 May 1998 11:50:55 -0400 >From: Paul Ferguson <ferguson(a)cisco.com> Hi Paul, and whoever might have gotten nervous about our (AS1853) "default-route": please calm down, there is no such beast, and we have never announced a default-route "at the Vienna Internet eXchange" (http://www.vix.at/) Please read below my answer/explanation to Paul Donner, the rumour is based on a misinterpretation ... As I'm not subscriber to the NANOG list, my CC to this list will most probably bounce back, therefore, Paul, please forward my answer there !!! Regards CP ======================================= Date: Wed, 06 May 1998 13:23:32 MET-DST From: "Christian Panigl, ACOnet/UniVie" <panigl(a)cc.univie.ac.at> To: pdonner(a)cisco.com CC: helpdesk(a)aco.net, panigl(a)cc.univie.ac.at Subject: VIX Looking Glass >Date: Wed, 06 May 1998 04:32:14 -0400 >From: "Paul G. Donner" <pdonner(a)cisco.com> >To: helpdesk(a)aco.net > >You guys are annoucing a default route at VIX. Is this what you really >want to be doing? I found this while using LOOKING GLASS. For example, >I was looking for a route to block 209.58.48.0/20. Now AS1583 does not >advertise a route for this block but they do offer to route any traffic >that is destined for this block. So you will be blackholing traffic >destined for this site from any of your peers. Dear Paul Donner, thanks for your hint. However, we are NOT announcing a default route "at VIX". AS1853 (ACOnet Backbone) is announcing a default route to AS1120 (ACOnet/VIX Service AS). The router Service.ACO.net/AS1120 is offering access to some special servers/services for all VIX members and is open to peer with all VIX members. Also the "VIX Looking Glass" is asking this router, which, besides of its VIX peerings, does not have any (not to speak about full) explicit routing information but a default route to the ACOnet Backbone (AS1853). This default route is NOT re-advertised to any VIX peer ! This looking glass wrt routing table therefore has only local meaning for VIX members and their customers. We might think of placing an explicit note there explaining the situation. Thanks for your input anyway ! Regards CP --- ---------------------------------------------------------------------- --- --- Christian Panigl : Vienna University Computer Center - ACOnet --- --- VUCC - ACOnet - VIX : -------------------------------------------- --- --- Universitaetsstrasse 7 : Mail: Panigl(a)CC.UniVie.ac.at (CP8-RIPE) --- --- A-1010 Vienna / Austria : Tel: +43 1 4277-14032 (Fax: -9140) --- --- ---------------------------------------------------------------------- ---

2 1

Routing Policy and http://rs.arin.net/ip-allocation.html
by Mike Tancsa 06 May '98

06 May '98

For the first time we have had to deal with Sprint's routing policy as defined by http://www.sprint.net/filter.htm. Here is the situation. One of our dialup customers wants to access his website in the 206.116.31.0/24 network at another provider. PSI is advertising it as a /24. According to Sprint's routing policy, they do not honour anything longer than a /19 in 206.0.0.0/8 . I am now in the position trying to explain to the very non-technical customer, why he cant reach that site, which from his point of view just seems like I am passing the buck... Anyways, my question is, are their any other ISPs/NSPs that follow the same guidelines. UUNet seems to honor it just fine, as does MCI. Is Sprint more hardnosed than most ? ---Mike ********************************************************************** Mike Tancsa (mike(a)sentex.net) * To do is to be -- Nietzsche Sentex Communications Corp, * To be is to do -- Sartre Cambridge, Ontario * Do be do be do -- Sinatra (http://www.sentex.net/~mdtancsa) *

9 8