March 2001 - Test - lists.nanog.org

Re: [NANOG] Re: new.net: yet another dns namespace overlay play
by Andrea Abrahamsen 08 Mar '01

08 Mar '01

I'm going to start a company and introduce a new TLD. .sucks. I wonder what ICANN would have to say about that. :) At 13:18 03/07/01, you wrote: Which brings me to the interesting borderline case: companyname.sucks.com Can WIPO take away hostname records?:) Pi

4 3

Performance at the limits
by Sean Donelan 08 Mar '01

08 Mar '01

The traditional telephone system has several well-known difficulties dealing with wide-spread emergencies. However, traditional reliability numbers don't tell the whole story. Emergency planners need to be careful when hearing that the telephone system is 99.999% reliable. The quoted figures exclude performance during periods of extreme stress. While other technologies may appear to have less reliability, you need to look at performance at the limits of the curve. Other technologies may have better performance characteristics at the extreme limits. Likewise, you need to be careful about applying solutions developed for the telephone system to other types of networks. http://www.nytimes.com/2001/03/08/technology/08TOWE.html While fiber optic technology proved invaluable, other technology fell short. The cellular telephones that controllers had come to rely on were useless after the earthquake, and land lines were jammed. The walkie- talkies they had kept after converting to cell phones were few and old. "We discovered that reverting back to basic two-way radios was what we really needed," Mr. Owen said. "If there is a next time, we'll have better two-way radio communications."

1 0

BGP deployment
by Ankur Goyal 08 Mar '01

08 Mar '01

Short time listener & first time caller. I have some questions regarding, the kind of configuration that is typically done, on a BGP-enabled router sitting at the edge of the network. My questions may have already been answered earlier, if that's the case, then someone please point me to the relevant mails... 1. In general, how many BGP peers should an edge router be capable of handling, more specifically, how many E-BGP peers and how many I-BGP peers are typically incident on one edge router. 2. In most situations the routes to the stub networks are manually injected into BGP. I feel that's the most common scenario but was wondering if running some IGP between the stub network and the provider edge router and then importing the routes from this IGP into BGP is also common ? Is this also done frequently enough ? Another option could be to run a BGP session between the stub network and provider edge router. Is this preffered ? Any help/suggestions would be highly appreciated. --Ankur

2 1

Re: new.net: yet another dns namespace overlay play
by Aaron Hopkins 08 Mar '01

08 Mar '01

-----BEGIN PGP SIGNED MESSAGE----- So the short answer minus the marketing-speak is that new.net is offering domains in new TLDs without having them in *.root-servers.net. The idea isn't new; people have been doing this for years with limited success. As domains in the new TLDs are obviously only useful if people can reach them, new.net is offering a variety of ways to be able to resolve them. Anyone in control of a recursive nameserver can: - - Slave a root zone with the new TLDs appended. - - Add stubs for the new TLDs pointing at udns[12].newdotnet.net. (I'm pleased with the UltraDNS infrastructure, which is serving similarly here to the gtld-servers.net but for new.net's TLDs. For those using the stub-zone hack, these are the only servers they ever see.) - - Replace the root cache with one pointing to ns[123].newdotnet.net, which will delegate as appropriate to the various TLD servers. (There are obviously too few of these available with not enough diversity, but more will be added soon.) The "launch partners" are listed as Earthlink, Excite@Home, NetZero. These shortly will be using one of the above to allow their users to reach the new TLDs. Given that the whole goal is for the domains to be resolvable, expect to see more. But these options aren't available to the average end user. Rather than setting up a few centralized recursive nameservers and trying to get every desktop in the world to send their traffic to them, new.net has elected to also make the domains reachable with ".new.net" appended. This allows for proper local caching and search-path related tricks. For all desktop users, the option is available to just add new.net to the search path. One of the functions of the Windows client offered on the site is a resolver plugin that appends ".new.net" on to some queries. It tries to cut down on the number of failed DNS requests that just would be generated by just adding new.net to the search path. It will also act as a sub-proxy for proxy users, appending ".new.net" on to some HTTP proxy requests. I'm not sure whats going on with a patent application on this. This is obviously less clean than just having ICANN list new TLDs. But given the ongoing frustration with ICANN, there is significant interest in a different approach, and new.net is taking one. I'm pretty sure that I didn't want it to come to this, and I'm not entirely convinced that anyone should be doing it. But personal reservations aside, its happening. And I intend to see that its done as well as possible. Aaron Hopkins Systems Engineer, idealab! Acting VP of Engineering, new.net -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.0.4 (GNU/Linux) Comment: For info see http://www.gnupg.org iQCVAwUBOqXvPEfJWHAEvsjBAQFgAQQAmtREC7ZCYLVrA2biCzriM1z07Jrk7oAc fG9znSGC7gtHrcsad2N7SR+wFsWP2v8I546zYMvpcfHDClRFyS11tohi85W/CfEQ 9kaWfe01Jzqri37v+jHfAyX7k7OrSz+9wRJoh72rTItMWgaLECA1wync//EJ7EYc o+zS2KTe2P4= =kJXu -----END PGP SIGNATURE-----

4 3

Re: new.net: yet another dns namespace overlay play (fwd)
by Shawn McMahon 07 Mar '01

07 Mar '01

On Wed, Mar 07, 2001 at 05:56:38PM -0500, Tuc wrote: > > > heimdall% whois verizonshouldspendmoretimefixingtheirnetworkandlessmoneyonlawyers.com > > Whois Server Version 1.3 > > Domain names in the .com, .net, and .org domains can now be registered > with many different competing registrars. Go to http://www.internic.net > for detailed information. > > No match for "VERIZONSHOULDSPENDMORETIMEFIXINGTHEIRNETWORKANDLESSMONEYONLAWYERS.COM". Ok, if you must; it wasn't taken away, they simply didn't choose to keep it. If you're using that as proof that they can't keep a domain with a trademark in the name: [smcmahon@qward smcmahon]$ whois verizonreallysucks.com [whois.crsnic.net] Whois Server Version 1.3 Domain names in the .com, .net, and .org domains can now be registered with many different competing registrars. Go to http://www.internic.net for detailed information. Domain Name: VERIZONREALLYSUCKS.COM Registrar: NETWORK SOLUTIONS, INC. Whois Server: whois.networksolutions.com Referral URL: www.networksolutions.com Name Server: FOO.NOTWORK.NET Name Server: PHALSE.2600.COM Updated Date: 04-apr-2000 >>> Last update of whois database: Wed, 7 Mar 2001 08:19:00 EST <<< The Registry database contains ONLY .COM, .NET, .ORG, .EDU domains and Registrars.

1 0

RE: new.net: yet another dns namespace overlay play
by brandon＠rd.bbc.co.uk 07 Mar '01

07 Mar '01

> You could have built a shared > registration system that could have encompassed all the non-ICANN TLDs, and > helped create something that would have really given serious challenge to > ICANN. With duplication it just adds weight to the ICANN "I told you so" arguments for not letting people decide this stuff for themselves "they'll only mess it up" and makes alternate root acceptance less likely brandon

1 0

Re: new.net: yet another dns namespace overlay play
by Richard A. Steenbergen 07 Mar '01

07 Mar '01

On Wed, Mar 07, 2001 at 05:36:13PM -0500, David R. Dick wrote: > > I wish more people would understand that just because > no one has implemented their idea before doesn't mean > that it wasn't thought of, then discarded as a bad idea. > > Then again, that sort of humility is probably not > prevalent at a company named "IdeaLab!" http://news.cnet.com/news/0-1007-200-5046211.html?tag=lh >From the company that brought us eToys.... 'nuf said... -- Richard A Steenbergen <ras(a)e-gerbil.net> http://www.e-gerbil.net/ras PGP Key ID: 0x138EA177 (67 29 D7 BC E8 18 3E DA B2 46 B3 D8 14 36 FE B6)

1 0

Re: new.net: yet another dns namespace overlay play
by Richard A. Steenbergen 07 Mar '01

07 Mar '01

On Wed, Mar 07, 2001 at 04:04:17PM -0500, Charles Sprickman wrote: > > On Wed, 7 Mar 2001, Patrick Greenwell wrote: > > > Parody and criticism are protected under trademark law(although those > > rights are being eroded quickly.) > > No one has a sense of humor any more. A lot of folks have had > <companyname>sucks.com taken away from them due to "trademark > violations". Among the many domains registered to "Bell Atlantic Trademark Services LLC": VERIZONSUCKS.COM VERIZONBLOWS.COM VERIZONBITES.COM VERIZONSHITS.COM VERIZONWIRELESSSUCKS.COM VERIZONWIRELESSBLOWS.COM VERIZONWIRELESSBITES.COM VERIZONWIRELESSBITES.NET VERIZON-TELEKOM.ORG etc etc etc... You could almost make a case that they're preventing people from making criticisms, but given Verizon's level of performance its entirely possible these are domains for a soon-to-be-launched new product line. -- Richard A Steenbergen <ras(a)e-gerbil.net> http://www.e-gerbil.net/ras PGP Key ID: 0x138EA177 (67 29 D7 BC E8 18 3E DA B2 46 B3 D8 14 36 FE B6)

1 0

GSR and bgp maximum-paths
by James DeMong 07 Mar '01

07 Mar '01

Long time listener, first time caller. I would like to enable multipath bgp on a GSR running 12.0(10)S that has two bgp neighbors in the same AS connected via OC-3 pos in order to load balance outbound traffic across the two pos connections. I've checked out the cisco bug search and haven't found any for this feature on this release of IOS. Is anyone else have any problems with "maximum-paths" on a GSR and if yes, what release of IOS? Any other issues I should be aware of? Thanks, James __ James DeMong Network Design Specialist TELUS Advanced Communications Phone: (403) 503-3718 Email: James.DeMong(a)telus.com

2 1

An example of reverse-hijack (was: new.net: yet another dns names pace overlay play)
by Doane, Andrew 07 Mar '01

07 Mar '01

Scott, I registered "aol.org" in 1995. It stood for "Andy On-Line" (my name). I used it for almost five years without any problems from America Online, mainly for email and to ftp back into my home network, etc. I did not have a website up. America Online sued and won a judgment without even my being present. They stated that it violated their trademark of "AOL", as they had used it first in commerce. They used the fact that I had no web site at www.aol.org to justify this claim, ignoring the fact that their was an MX record and some other A records. A few "interesting facts": 1. In 1995, even if America On-Line wanted to registered .org they couldn't. Back then the rules were followed, and .org was for "private organizations and individuals". I registered it as an individual for non-commerce use. So their argument that they used it in commerce first is moot - as a .org domain you weren't supposed to. 2. America Online just recently acquired the right to protect "AOL" from the courts. Traditionally speaking, initials cannot be trademarked. If a company reaches "household word" status then they can get protection. IBM and GM would be examples. In 1995 when I registered aol.org, America Online did NOT have this protection. 3. Network solutions suspended the domain without warning by request of America On-Line. My phone calls to network solutions resulted in "call this person", which turned out to be America On-Line's attorneys. NS did this before America Online had received a judgment from the court. 4. Network Solutions and America Online claim they tried to reach me. They didn't. The address on WHOIS was wrong, however the email addresses and phone numbers were valid. Very convenient for them. However, the automatically generated renewal notice that I needed to pay my $35/year fee from NS made it to me. Funny, eh? 5. I hired an attorney to fight it, not because I wanted cash out on America Online (in fact, I didn't ask for this - I just wanted my personal domain back that I had been using for almost 5 years without issue). The net result was it would have cost me a mint to fight it against a multi-billion dollar company with endless resources. I caved and just let them have it. To date, they have not used it. They blackholed the domain. Someone explain the point in that. Obviously they felt people could confuse "aol.com" and "aol.org". So much so that they pointed www.aol.org to www.aol.com? No. So much that they put in MX records so email accidentally sent to joeuser(a)aol.org got delivered or at least a response back to the person who sent the email that they have the email address wrong? No. Big business wins against the individual by manipulating our government to draft legislation in their favor and then using it after the fact. There's your example. -Andrew NOT andy(a)aol.org -----Original Message----- From: Scott Gifford [mailto:sgifford@tir.com] Sent: Wednesday, March 07, 2001 3:16 AM To: William Allen Simpson Cc: nanog(a)merit.edu Subject: Re: new.net: yet another dns namespace overlay play William Allen Simpson <wsimpson(a)greendragon.com> writes: > Patrick Greenwell wrote: > > > > On Tue, 6 Mar 2001, Paul A Vixie wrote: > > > ICANN's prospective failure is evidently in the mind of the beholder. > > > > Besides producing a UDRP that allows trademark interests to convienently > > reverse-hijack domains > > Awhile back, somebody made a similar accusation. So, I spent the > better part of a weekend reviewing a selection of UDRP decisions. > Quite frankly, I didn't find a single one that seemed badly reasoned. > > Could someone point to a "reverse-hijacked" domain decision? Assuming that I'm correctly understanding what is meant by "reverse-hijacked", the most notorious case I'm aware of is "walmartsucks.com". This domain was taken from an owner serving up criticism of Wal-Mart, and given to Wal-Mart. Wal-Mart apparently claimed that this domain name was so similar to their actual trademark, customers could be confused into visiting the wrong site, and ICANN somehow agreed. I don't know where the official ICANN ruling is on this, but I recall seeing it discussed in a number of places at the time. Let me know if you can't find a reference, and I'll see if I can dig one up. -----ScottG.

2 1