- NANOG - lists.nanog.org

Re: UUNET is not the Internet (and neither is AOL)
by Dave Israel 07 Oct '02

07 Oct '02

On 10/5/2002 at 12:30:36 +0000, Tim Thorne said: > After reading all the stories about what supposedly happened does > anyone know what really happened? Did UUNet US really do an IOS > upgrade on a sizable proportion of their border routers in one go? > This seems like suicide to me. What possible reason could there be for > a network-wide roll out of an untested IOS apart from being in the > mire already? The assumption that it was untested is probably an unfair one. Once a network reaches a certain size, it is very difficult to simulate it in a lab. Number of routes/updates, variety of packet destinations, different card revisions and layouts... heck, even statistically, you have problems. An issue that appears 5% of the time will only show up in a a 10-router test lab half the time, but in a 400-router network it'll pop up on about 20 routers and wreck your whole day. And when you're out of cash, you can't really afford to devote lots of hardware to a lab. I'm not saying that their testing procedures were correct, or that they tested the image as well as they could have... but the assumption that, if it blew up in the field it must not have been tested at all probably isn't accurate. -Dave

2 1

Re: UUNET is not the Internet (and neither is AOL)
by Vicky O. Mair 07 Oct '02

07 Oct '02

Hi there, What really confuses the heck out of me is that a company this size can't control/monitor their change management??. Then again not having all the facts has had everyone perplexed. later, vicky At 07:38 PM 10/5/2002 -0400, you wrote: >On Sat, 5 Oct 2002, Tim Thorne wrote: > > After reading all the stories about what supposedly happened does > > anyone know what really happened? Did UUNet US really do an IOS > > upgrade on a sizable proportion of their border routers in one go? > > This seems like suicide to me. What possible reason could there be for > > a network-wide roll out of an untested IOS apart from being in the > > mire already? > >Corporate culture is the hardest thing to change in a company. You'll need >to talk with your Worldcom account rep about what happened, and what >Worldcom intends to do about it. In the past, Worldcom has not been very >open or transparent when it has had network problems.

6 5

RE: Good quotes on importance of good network addressing
by Irwin Lazar 07 Oct '02

07 Oct '02

-----Original Message----- From: Scott Francis [mailto:darkuncle@darkuncle.net] Sent: Monday, October 07, 2002 10:56 AM To: Stephane Bortzmeyer Cc: nanog(a)merit.edu Subject: Re: Good quotes on importance of good network addressing On Mon, Oct 07, 2002 at 09:49:28AM +0200, bortzmeyer(a)gitoyen.net said: > On Thursday 3 October 2002, at 12 h 23, > Scott Francis <darkuncle(a)darkuncle.net> wrote: > > > Not sure how applicable it may be, but the OpenBSD FAQ has referenced (since > > at least 2.7) a paper called "Understanding IP Addressing" that I found to = > > be > > pretty useful. > > http://www.3com.com/corpinfo/en_US/technology/tech_paper.jsp?DOC_ID=3D135 > > http://www.3com.com/other/pdfs/infra/corpinfo/en_US/501302.pdf > > It seems quite old and not very practical : > > * it never mentions RFC 1219, > * it mentions IPv6 in a few words, without any practical considerations, > * it explains the old classfull addressing first, instead of talking CIDR > right from the beginning. have a look at Designing Addressing Architectures for Routing and Switching by Howard Berkowitz. You'll find analogies between IP addressing and the Ming dynasty. Irwin

1 0

redistribute bgp considered harmful
by Sean Donelan 07 Oct '02

07 Oct '02

Should the Service Provider version of routing software include the redistribute bgp command? Other than CCIE labs, I haven't seen a real-world use for redistributing the BGP route table into any IGP. If the command was removed (or included a Are your sure? question) what would the affect be on ISPs, other than improving reliability by stopping network engineers from fubaring a backbone?

8 8

RE: iBGP next hop and multi-access media
by Martin, Christian 07 Oct '02

07 Oct '02

Ralph, What you are asking for is dynamic routing protocol to be able to tell you how to do something locally scoped. This is why we have redirects on ethernets. But, since you must do this with BGP, try this... (I haven't) router bgp x neighbor 10.10.10.2 remote-as x neighbor 10.10.10.2 route-map foofoo in ! route-map foofoo permit 1 match ip address 1 set interface e1 ! access-list 1 permit 172.16.16.0 -chris > -----Original Message----- > From: Ralph Doncaster [mailto:ralph@istop.com] > Sent: Sunday, October 06, 2002 9:06 PM > To: E.B. Dreger > Cc: nanog(a)merit.edu > Subject: Re: iBGP next hop and multi-access media > > > > > RD> When I setup a situation like the above, with Router B > advertising > > RD> the 172.16.16.0/24 to router A, router A sees a next hop of > > RD> 10.10.10.2. This is not good since packets from A going to the > > RD> 172.16.16 subnet get sent to Router B, which then ARPs the > > RD> desitnation, instead of just being ARPed by router A. > > > > Is this what you're trying to do: > > > > route-map <foo> > > match <whatever> > > set ip next-hop <something> > > Not really, what I want is router A to learn that ther is no > next hop IP- the subnet is on the local ethernet. > > -Ralph >

1 0

RE: UUNET Routing issues
by Brennan_Murphy＠NAI.com 07 Oct '02

07 Oct '02

Are the tickets closed yet? I'm tempted to call in and see if I can get a grasp of the scope and nature of the problem. But maybe it would be best if someone simply posted a brief summary of what is publicly known about the issue....to be followed by resonable speculation peppered with some wild speculation. Seems to be a relevant topic of mild interest to the majority of the list if in fact the problem is impact multiple UUNET locations. -----Original Message----- From: Eric Whitehill [mailto:eric@botbay.net] Sent: Thursday, October 03, 2002 10:19 AM To: Patrick_McAllister(a)WASHGAS.COM Cc: Manolo Hernandez; Nanog; owner-nanog(a)merit.edu Subject: Re: UUNET Routing issues For T-1 customers, the master Ticket Number is 651744 For customers with DS/OC gear, that master ticket number is 651751. I came to this information after calling their noc and asking. :) -Eric

15 22

RE: iBGP next hop and multi-access media
by Charles Youse 07 Oct '02

07 Oct '02

Really, the only way this could happen is if Router B is not announcing its routes to 172.16.16/24 and Router A has a default route to its Ethernet interface. C. -----Original Message----- From: Ralph Doncaster [mailto:ralph@istop.com] Sent: Sunday, October 06, 2002 9:06 PM To: E.B. Dreger Cc: nanog(a)merit.edu Subject: Re: iBGP next hop and multi-access media > RD> When I setup a situation like the above, with Router B advertising > RD> the 172.16.16.0/24 to router A, router A sees a next hop of > RD> 10.10.10.2. This is not good since packets from A going to the > RD> 172.16.16 subnet get sent to Router B, which then ARPs the > RD> desitnation, instead of just being ARPed by router A. > > Is this what you're trying to do: > > route-map <foo> > match <whatever> > set ip next-hop <something> Not really, what I want is router A to learn that ther is no next hop IP- the subnet is on the local ethernet. -Ralph

1 0

Telco cages?
by Alex Rubenstein 06 Oct '02

06 Oct '02

I am looking for a manufacturer of telco cages used in datacenter applications; any pointers would be appreciated. -- Alex Rubenstein, AR97, K2AHR, alex(a)nac.net, latency, Al Reuben -- -- Net Access Corporation, 800-NET-ME-36, http://www.nac.net --

4 3

RE: Thanks! Re: Re: Root DNS Server Issues?
by Doug Barton 05 Oct '02

05 Oct '02

On Fri, 4 Oct 2002, Bodie Francis wrote: > I am sure Doug and others could use this to reduce the typing (the reason > Unix was invented) when doing recursive queries for delegations (it even > identifies lame delegations quite nicely.) Sorry if I wasn't clear. My example was intended to be pedantic, as opposed to actually being useful. dnstracer is an interesting program, but I have a home-grown tool of my own that I use for day to day delegation chain sniffing. The dig in bind 9 also has a new trace option that's fairly useful. Doug

1 0

RE: redistribute bgp considered harmful
by Andrew.McConnell＠sungard.com 04 Oct '02

04 Oct '02

Let's not forget about the use of VRF's. BGP is not used exclusively for sending public routes in our network. Just a thought. -- Andrew McConnell Network Implementation Engineer SunGard Network Technologies 401 North Broad St Philadelphia, PA 19108 Charles Youse <cyouse@register. To: 'Sean Donelan' <sean(a)donelan.com>, nanog(a)merit.edu com> cc: Sent by: Subject: RE: redistribute bgp considered harmful owner-nanog@merit .edu 10/04/2002 06:09 PM I've never subscribed to the "Are you sure?" concept, or preventing problems by removing functionality, effectively tying an operator's hands behind his/her back. The fact is that redistributing BGP into an IGP can have its uses (though not usually, okay, never, when carrying a full table on the public Internet) and I'd hate to see the messy workarounds that would come about, when the solution could otherwise be straightforward. My Windows workstation asks me "Are you sure?" all the time. Just annoying, really. C. -----Original Message----- From: Sean Donelan [mailto:sean@donelan.com] Sent: Friday, October 04, 2002 6:01 PM To: nanog(a)merit.edu Subject: redistribute bgp considered harmful Should the Service Provider version of routing software include the redistribute bgp command? Other than CCIE labs, I haven't seen a real-world use for redistributing the BGP route table into any IGP. If the command was removed (or included a Are your sure? question) what would the affect be on ISPs, other than improving reliability by stopping network engineers from fubaring a backbone?

1 0