- Test - lists.nanog.org

Fiber Cut
by scott w 13 Nov '98

13 Nov '98

290 DS-3s cut in Mount Green, Wyoming. Backhoe at site. *heh* let's don't go there. Got those t-shirts yet? Tech's currently splicing fiber and rerouting as they go. All this from MFS, so take it with what you will... scott

1 0

Re: Journal of Internet Disasters
by Paul Vixie 13 Nov '98

13 Nov '98

> This should go on the name-droppers list, but here goes.... these days it's not clear whether namedroppers is an operations list or a protocol list or still both. i think nanog is a fine forum for this: > What do we know about the events with the name servers > > - f.root-servers.net was not able to transfer a copy of some of > the zone files from a.root-servers.net > - f.root-servers.net became lame for some zones just COM. > - tcpdump showed odd AXFR from a.root-servers.net just a lot of missed/retransmitted ACKs. > - [fjk].gtld-servers.net have been reported answering NXDOMAIN to > some valid domains, NSI denies any problem the nanog archives include some dig results that are hard for NSI to deny. > Other events which may or may not have been related > - BGP routing bug disrupted connectivity for some backbones in the > preceeding days this turned up a performance problem in BIND's retry code, btw, but was not otherwise related to the COM lossage of yesterday (as far as i know). > - Last month the .GOV domain was missing on a.root-servers.net due > to a 'known bug' affecting zone transfers from GOV-NIC different bug. that one causes truncated zone transfers; the secondary zone files on [fjk].gtld-servers.net yesterday were not truncated and it just took a restart to make them stop behaving badly. > - Someone has been probing DNS ports for an unknown reason > > Things I don't know > - f.root-servers.net and NSI's servers reacted differently. What > are the differences between them (BIND versions, in-house source > code changes, operating systems/run-time libraries/compilers) they are completely different systems (solaris vs. digital unix) running the same (unmodified) bind 8.1.2 sources, which had completely different failure modes for completely different reasons. > - how long were servers unable to transfer the zone? The SOA says > a zone is good for 7 days. Why they expire/corrupt the old zone > before getting a new copy? damn good question. i'll look into that. shouldn't've happened. > - Routing between ISC and NSI for the preceeding period before the > problem was discovered there was asymmetry (they reached me via bbnplanet, i reached them via alternet). they are now preferring alternet to reach me, so we have better path symmetry now. but their first mile is still congested and i am still retransmitting a lot of ACKs. > Theories > - Network connectivity was insufficient between NSI and ISC for long > enough the zones timed out (why were other servers affected?) other servers are more conservative, and had switched to manual daily FTP of the COM zone longer ago than F has done. (with manual daily FTP you get the advantages of gzip, and of the pretense of "zone master" status while you manually retry after timeouts. AXFR needs those properties.) > - Bug in BIND (or an in-house modified version) (why did vixie's and > NSI's servers return different responses?) there's definitely a bug in BIND if [fjk].gtld-servers.net were able to return different answers after restarts with no new zone transfers. (i'm sitting here wishing i had core dumps.) > - Bug in a support system (O/S, RTL, Compiler, etc) or its installation > - Operator error (erroneous reports of failure) > - Other malicious activity? i think there were a goodly number of procedural errors. -- Paul Vixie <paul(a)vix.com>

1 0

Watch found at NANOG 14
by Pam Ciesla 13 Nov '98

13 Nov '98

There was a watch found at the NANOG 14 Meeting in the Imperial Ballroom. If you have lost a watch and think this might be yours please e-mail me with description and mailing address. Thanks Pam Pam Ciesla Merit Network,Inc. 734-936-0172 - phone 4251 Plymouth Road 734-647-3185 Bldg.I, Suite C pam(a)merit.edu Ann Arbor, MI 48107

1 0

logical subnet overlaying...
by Kyle D. Smith 13 Nov '98

13 Nov '98

hello. I'm not sure if this is the right mailing list to ask, but I hope that someone could possiably help.. I have a network with a cisco 2501 router running IOS 10.2(9), and we have a block of addresses from our internet service provider (60 addresses, 206.206.162.128, start. with 1 for network, and 1 for brodcast) we recently requested another block of 32 from our ISP, and they gave us the following: 207.66.81.96, start...with 1 for network/1 for brodcast. I asked a similar question on a mailing list before (not sure if it was this one) and was told to login to the router, then: router> en pass:..... router# conf term router <conf> int e 0 router <subif> ip address existing_ip existing_subnetmask secondary crtl-z write i tryed this, but it did not work. I also tryed replacing the exiting ip/subnet with the new one... but that did not work either... am i doing something wrong? or do i just need a 11.xx(xx) IOS ? Any help would be greatly apreciated. thanks. =-=-=-=-=-= -Kyle Donald Smith -Systems Administrator -Community Internet Access, Inc. -peaches(a)cia-g.com

2 1

Re: gotta love those internic boyz (fwd)
by Paul Vixie 13 Nov '98

13 Nov '98

At 12:05 13/11/98 +1100, somebody whose name wasn't quoted wrote: >Oh boy, seems Vixie shoudl join IRSC, we don't have these kinds of problems. > >Perhaps it's time the various parties that wish to stand by the history >realised that cooperation on a wider scale might help? Cooperation occurred and is occurring. "IRSC" is semantic line noise. at(a)ah.net (Adam Todd Snr) added: > Hey Paul Vixie! Attention Paul Vixie! Yo. > Now that there have been a few too many stuff ups with the old legacy Root > Servers, and especially as you've copped the heat this time, why not offer > the option and ship bind with an IRSC section! Because DNS was designed to replace HOSTS.TXT, and BIND is a DNS implementation. If you want every site's "view of the naming universe" to depend on the politics of that site's owners, then you should avoid DNS and use something which will more reliably suit your private naming purposes, such as HOSTS.TXT. The 4BSD tapes have a useful utility which can transform files in HOSTS.TXT format into /etc/hosts format. Go to it! > Then let the consumers choose! See above. DNS is a distributed, reliable, coherent, autonomous database. You mess with any of those things and you have something which isn't DNS or which is a horrid example of DNS. Specifically, please stop pretending that coherence is optional. > Wouldn't it bet better they had the choice to accept a service, then be > DENIED the knowledge and option of selecting a service. Seems like a lot of folks keep blathering the same old crap about how the proletariat ought to control the naming system, by chaos if nothing else, and that your views are adequately represented. BIND, being a DNS implementation, takes no notice of your activities, which are not DNS at all. -- Paul Vixie <paul(a)vix.com>

1 0

Re: gotta love those internic boyz (fwd)
by Adam Todd Snr 13 Nov '98

13 Nov '98

At 12:05 13/11/98 +1100, you wrote: >Oh boy, seems Vixie shoudl join IRSC, we don't have these kinds of problems. > >Perhaps it's time the various parties that wish to stand by the history >realised that cooperation on a wider scale might help? Hey Paul Vixie! Attention Paul Vixie! Now that there have been a few too many stuff ups with the old legacy Root Servers, and especially as you've copped the heat this time, why not offer the option and ship bind with an IRSC section! Then let the consumers choose! Wouldn't it bet better they had the choice to accept a service, then be DENIED the knowledge and option of selecting a service. THEY STOLE OUR BABY AJ WITHOUT TELLING US WHY http://www.aj.inau.com/docs -------------------------------------------------------------------------- Adam Todd http://adamtodd.ah.net Business Development, Technology, Domain Registration and Network Advisory Phone +61 2 9729 0565 - Todd Corporation http://www.todd.inoz.com Fax +61 2 4659 6786 - AHNET http://www.ah.net - AURSC http://www.aursc.ah.net Telstra Reseller and Telstra Convey Member (Not an Employee of Telstra) Get the DOMAIN NAME HANDBOOK NOW http://www.domainhandbook.com

1 0

Internic, et al.
by Tim Wolfe 13 Nov '98

13 Nov '98

Just a thought. The next time gtld servers start puking and Nic starts playing "we'll get around to fixing it in the next server update" why don't all the ISPs start having their users send all of their complaints to complaints(a)internic.net Won't help much but it sure would feel good... Tim ---------------------------------------------------- Timothy M. Wolfe | Why surf when you can Sail? tim(a)clipper.net | Join Oregon's Premier Sr. Network Engineer | Wireless Internet Provider! ClipperNet Corporation | http://www.clipper.net/ ----------------------------------------------------

2 1

summary: gateway failover.
by foo 13 Nov '98

13 Nov '98

first of all, sorry for the duplicate message. i originally sent it without my proper from address to avoid having my primary mailbox spammed, but thought it wouldnt go through because the from address wasnt on nanog-post... second, i incorrectly referred to the rfc for IRDP as 1285, when it is 1256. re: IRDP support. the following os's/routing daemons have support: gated (version ?) Solaris (in.rdisc(1)) Windows NT (with sp3) Windows 95 (with Sockets 2) a couple of useful urls were sent my way, the most interesting being: http://www.eddieware.org which details an Open Source HA project by Ericsson (very cool), which is unfortunately only supported under Linux for the time being, though Solaris/FreeBSD support is slated for December. another from a usenix/lisa presentation with some good info is here: http://dilbert.dcs.it.mtu.edu/failover/ also, many suggestions were given for ways to accomplish this goal with network layer stuff (ospf, rip, interface aliases, proxy arp, etc.), though i was looking for an application layer solution, primarily. if you'd like details on the various suggestions, email me privately. -foo

1 0

Re: Is it time for an disruption analysis working group for the Internet?
by Sean Donelan 12 Nov '98

12 Nov '98

pete(a)kruckenberg.COM (Pete Kruckenberg) writes: >How is this handled in other networked industries? I'm sure that the same >issues of proprietary information and public humliation exist there; how >do they deal with it? Other industries, e.g. electric, airline, nuclear, telephone, railroad, banking, etc, tend to have two different reporting regimes. One for 'major events,' and one for 'near misses.' For major events, the issues of proprietary information and public humiliation tend to be less. Mainly because everyone already knows you had a major event (e.g. three-mile island, at&t nationwide failure) and the company's and industry's goals tends to focus on restoring public confidence. This seems to be true no matter what the root cause of the problem. After a major event, company records tend to get opened either voluntarily or involuntarily anyway. So most companies prefer the P.R. of saying they are cooperating fully with investigators. Although as NASA found out, sometimes you get a Richard P. Feynman on the board, who doesn't always follow the company line. I think, for some definition of major events, it should be possible to get cooperation from a large number of service providers. Once the event has happend, it is usually in their self-interest to participate in the process. Of course, there will be a few service providers who won't participate on principle or because of internal disorganization. But that's what the press is for.... For near-misses, the reporting parties tend to be individuals (a.k.a. whistleblowers) and confidentiality is more important due to the retribution even accurate information receives. And in general, 'near misses' don't attract much attention because by definition, the crisis didn't happen. I think any type of near miss reporting and analysis will be very difficult. I could rant about the U.S. Government's plans in this area are, but that is a different mailing list. -- Sean Donelan, Data Research Associates, Inc, St. Louis, MO Affiliation given for identification not representation

1 0

Re: ARIN?
by Andrea Di Lecce 12 Nov '98

12 Nov '98

At 10:40 11/5/98 -0500, Brandon Ross wrote: >> Looks alot like /14 & change. That would be 256K plus delegated - 210K users >> Hey, your right, not a bad packing ratio. Hows yours? :) > >How's 2 /16's & change, 128k addresses - 455k users. I think I win. (In >an effort to fully disclose, about 20% of those users are on wholesale >dialup providers). Cable modem subscribers normally are not dynamically assigned IPs like dialups. Besides, they are advertised as a constant connection, and a lot of people use it as such. ************************************************** Andrea Di Lecce (416) 935-5700 Rogers@Home Network Operations 1 Mount Pleasant Road, Toronto, M4Y 2Y5 **************************************************

26 39