June 2002 - Test - lists.nanog.org

Internet Routing Registry Help Desk at NANOG 25
by Larry J. Blunk 05 Jun '02

05 Jun '02

Merit RADB technical staff will be hosting an Internet Routing Registry (IRR) Help Desk at NANOG 25. The desk will be staffed on Sunday 7:00 - 9:30PM, and from 1:00 - 1:30PM and during breaks on Monday and Tuesday. A list of potential discussion topics are included below -- -- General IRR topics * RPSL concepts and usage * Benefits of using the IRR * Using PGP and CRYPT-PW authentication * Using the IRR to configure routers -- RADB/IRRd specific topics * Registering for RADB service * Updating/Removing your RADB objects for consistency * Determining the status of your RADB registration and objects * RADB billing questions and issues * Status of the RADB clean-up * Feedback on the RADB service * Demo of upcoming RADB web interface for updates * IRRd software questions and feedback In addition, RIPE NCC technical staff will be available Sunday from 7:00 - 9:30 PM to accept feedback and questions on the IRRToolSet. For further information or questions, please contact db-admin(a)merit.edu. -Larry J. Blunk Merit

1 0

Re: Automated Reply: OT: Re: Bogon list
by Scott Francis 05 Jun '02

05 Jun '02

On Wed, Jun 05, 2002 at 03:18:58PM -0400, MSAutoReply(a)morganstanley.com said: > This is an auto-generated system message. Please do not reply to this > address. [snip legalese] Whoever this is, will you PLEASE fix your auto-noise generator to not pollute mailing lists? Apologies to the list for the (twice-) wasted bandwidth; I have no other way to contact the person in question. Yes, I realize this just generated another auto-reply. *sigh* -- Scott Francis darkuncle@ [home:] d a r k u n c l e . n e t Systems/Network Manager sfrancis@ [work:] t o n o s . c o m GPG public key 0xCB33CCA7 illum oportet crescere me autem minui

1 0

Re: KPNQwest ns.eu.net server.
by John L Crain 05 Jun '02

05 Jun '02

Very welcome, Thanks for this Dani it gives everyone much more breathing space. I think this a fantastic effort on behalf of the NCC, the folks at eu.net and anyone else involved, they should all be applauded for it. A few logistical questions that may be relevant to those affected. Is there a projected time when this service may be discontinued? I can imagine that the RIPE NCC doesn't want to indefinitely run secondary for end sites? If not: This is in the same AS as ns.ripe.net. Is it also physically on the same infrastructure? What I guess I'm asking is: To what extent do those using both ns.ripe.net and ns.eu.net have redundancy between the two servers? JC At 07:25 PM 6/5/2002 +0200, Daniel Diaz wrote: >Dear all, > > >Given the current situation of KPNQwest and the possibility >of its services going offline sometime soon, the RIPE NCC in >agreement with KPNQwest will be temporally hosting this >server (ns.eu.net) in its premises. > >This is to avoid major problems in the Internet as this server >is secondary for a large number of ccTLD's zones, and thousand >other zones. > >We (AS3333) will be soon announcing the 192.16.202.0/24 prefix. > > >Trusting that this is welcome news. > >Best regards. > >-- >RIPE NCC > > >-- >Daniel.Diaz >Operations Manager >RIPE NCC

1 0

Route Authentication Request Results
by Barbara Fraser 05 Jun '02

05 Jun '02

Thanks to all the people who responded to your question about the use of BGP-MD5 and/or OSPF/IS_IS MD5. I'm collating the responses today and will send a summary to the list. Barb Barbara Fraser Consulting Engineer Cisco Systems, Inc. Phone: +1 (408) 525-1735

1 0

Re: Bogon list
by Miquel van Smoorenburg 05 Jun '02

05 Jun '02

In article <cistron.20020605144803.GB12164(a)overlord.e-gerbil.net>, Richard A Steenbergen <ras(a)e-gerbil.net> wrote: >On Wed, Jun 05, 2002 at 08:34:58AM +0000, Miquel van Smoorenburg wrote: >> >> I haven't seen a 'icmp source lo0' interface command yet. Hopefully >> it will be added for ipv6 so exchanges can use link-local addressing >> (ipv6 has no fragmentation, PMTUd is mandatory). > >I'm not terribly sure why you would want to make traceroutes lose all >information about the circuits you're traveling through. It would make >diagnostics an everloving nightmare, IMHO. With link-local addressing and ip verify unicast reverse you're not going to see any TTL exceeded traceroute packets from link-local addresses anyway. And no ICMP size exceeded packets either, which was my point. But indeed it should be 'ip icmp type <TYPE> source <IP|INT>', with at least a default 'ip icmp type 4 source lo0' or similar for interfaces with link-local addressing. Mike.

1 0

Re: Bogon list
by Miquel van Smoorenburg 05 Jun '02

05 Jun '02

In article <cistron.20020605013323.A7C1EC7936(a)cesium.clock.org>, Sean M. Doran <smd(a)clock.org> wrote: > >| Why treat exchange subnets differently to any other bit of backbone >| infrastructure? > >Oh, I wholeheartedly agree. I would love them all to use RFC 1918 >addresses, because it is VERY VERY VERY rare that anything outside >the scope in which the 1918 local use addresses are unique actually >has to communicate with backbone infrastructure of any type. And again Path MTU discovery gets broken. >What communication can your workstation have with an XYZNET router? Receive ICMP size exceeded packets ? I haven't seen a 'icmp source lo0' interface command yet. Hopefully it will be added for ipv6 so exchanges can use link-local addressing (ipv6 has no fragmentation, PMTUd is mandatory). Mike.

3 2

Meltdown somewhere?
by John Palmer 05 Jun '02

05 Jun '02

Is there some sort of a router meltdown somewhere this evening? I cant get through to most destinations, either with TELNET, http, ftp, DNS - nothing.

4 3

Its NAS again
by John Palmer 05 Jun '02

05 Jun '02

Update - it appears to be either NAS or COVAD that has melted down on the East coast. It seems to be affecting here as well. They even have a message on their customer service line.

2 1

Re: Bogon list
by smd＠clock.org 04 Jun '02

04 Jun '02

| Tweaking our Looking Glass software by itself would not fix the problem | (ours doesn't have this problem anyway). To fix the problem everyone | would have to tweak their Looking Glass software since the problem can | be seen when someone traceroutes from a peer or 3rd party's Looking | Glass into our customer (in the event they weren't receiving the IXP | blocks from us). You sort-of want to convince nth parties to launch their traceroutes _from_ your looking glass, rather than from random places. Targeting people who look up in-addr.arpa mappings, you could always emit pointers to would-be tracerouters -- get yer real data at http://... Points to the person who first puts such a thing into the DNS. | One better might be to have the Looking Glass participating routers | manipulate their source IP address for pings and traceroutes. Yes, this is a good idea. Sean.

2 1

Re: Bogon list
by smd＠clock.org 04 Jun '02

04 Jun '02

| Traceroute to www.foo.com, see it goes through an exchange. Ping | the router on the far end of the exchange, "host unreachable". How do you know the name of the thing at the far end of the exchange? Is it traceroute? Did the traceroute come from a looking glass, perhaps, or pass through a router controlled by the network participating in the exchange? If someone massaged or annotated the results of the traceroute, do you think you might a/ not ping that address, b/ ping an address which somehow corresponds to that router, c/ not know what the hell is going on at all -- maybe there's an MPLS tunnel across 15 routers! | Traceroute to it, "host unreachable" at the first default free | router. Not only will this confuse your customers, but often first | level support staff. The latter can be programmed to use an expert system that looks alot like a looking glass, but returns edited or annotated traceroutes instead. Hey, an in-house FAQ with hyperlinks for your support staff, that's too novel! Let's leak prefixes instead. | Also, you buy service from your favorite default free network. | They aren't present at an exchange. You want to traceroute/ping | a host on that lan. You can't, unless someone tranists the exchange. : macosx ; host some-router-at-mae-xyzzy.foo-bar.net <some loopback address> : macosx ; host some-router-at-mae-xyzzy-actual-interface-address.foo-bar.net <some RFC 1918 address> Which do you more likely ping? This is done all over the place now, unfortunately minus the 1918 addressing. Are there router implementations out there which ALWAYS do the wrong thing, and use the actual-interface-address when sending unreachables & exceededs? (Well, yes, I know there are. Sigh.) Sean.

1 0