- NANOG - lists.nanog.org

Crazy flying netbios packets
by Pete Ashdown 11 Sep '98

11 Sep '98

Has anyone done any estimates on how much net-wide traffic is useless netbios udp? Are there any suggestions for cutting large chunks of this out of my network without punishing SAMBA and other users who need it?

5 4

Mirroring web sites
by Sean Donelan 11 Sep '98

11 Sep '98

Thanks for the offers of mirroring. In case anyone didn't notice, there are some layer 8 & 9 issues going on. All the 'official' mirror sites are currently under the control of the legislative branch of government. The seperation of powers comes to the Internet. I have no inside information, but expect the response to be posted on an executive web site, e.g. www.whitehouse.gov. Hopefully by tomorrow afternoon there should be about 5 official sites. At the present time here are the mirror sites up and running: InternetMCI http://www.house.gov/icreport DRA Net http://thomas.loc.gov/icreport GTEI/BBN http://www.access.gpo.gov/congress/icreport Media outlets looking for 'favored' access should contact the public affairs office of either the House, the Library of Congress, or the Government Printing Office. Access is controlled by the House of Representatives. I would ask anyone who has set up a robot to automatically check the web sites for updated pages, and try to copy them, to make sure they are well-behaved. Ill-behaved robots may be mistaken for a denial of service attack, and action may be taken to block them. I don't know what InternetMCI or GTE have prepared. -- Sean Donelan, Data Research Associates, Inc, St. Louis, MO Affiliation given for identification not representation

1 0

Re: NSI Bulletin 098-010 | Update on Whois
by Jim Fleming 11 Sep '98

11 Sep '98

-----Original Message----- From: Leo Bicknell <bicknell(a)dimension.net> Newsgroups: maillist.nanog To: nanog(a)merit.edu <nanog(a)merit.edu> Date: Wednesday, September 09, 1998 4:53 PM Subject: Re: NSI Bulletin 098-010 | Update on Whois >In article <Pine.BSD/.3.91.980909153738.7931A(a)crispy.iconn.net> you write: >>According to David Holtzman of NSI (i asked him), the restrictions on whois >>are merely for technical reasons. I believe that he believes this, and >>from his point of view they have every right to filter/limit obnoxious or >>badly configured hosts/sites. > > I have a call to arms. If we could get orginizations to >mirror the whois data, and provide full, public access to it >(via the current whois database) we could remove some of the >dependency on NSI. After all, they don't run the only root, lots >of other people run them as well...why should they run the only >whois server? > As far as I know, the rwhois interface is more oriented toward the approach that you propose. You can walk the rwhois protocol thru a query or two via a simple telnet interface. The network of rwhois servers transfer complete zones. I think the original idea was to make it more like the DNS, only with contact information. Here is a random query about dimension.net. Note the minus sign to indicate commands. The "-holdconnect on" command keeps the link open for multiple queries. It does not take a rocket scientist to figure out from the rwhois protocol how to transfer a large number of entries. Note, there are 5,532,210 "objects". $ telnet rwhois.internic.net 4321 Trying 198.41.1.6... Connected to root.rwhois.net. Escape character is '^]'. %rwhois V-1.5:003eff:00 root.rwhois.net (by Network Solutions, Inc. V-1.5.2rri-d l) -status %status limit:20 %status holdconnect:OFF %status forward:OFF %status objects:5532210 %status display:DUMP %status contact:rwhois-bugs@rwhois.net %ok -holdconnect on %ok dimension.net domain:Class-Name:domain domain:Auth-Area:net domain:ID:DIMENSION2-DOM.net domain:Handle:DIMENSION2-DOM domain:Domain-Name:DIMENSION.NET domain:Org-Name:Dimension Enterprises domain:Street-Address:1039 Sterling Rd, Suite 201 domain:City:Herndon domain:State:VA domain:Postal-Code:22170 domain:Country-Code:US domain:Server;I:GIZMO2-HST.NET domain:Server;I:USSEN-HST.ORG domain:Server;I:PUMA12-HST.COM domain:Admin-Contact;I:FM136.NET domain:Tech-Contact;I:JTB2.NET domain:Billing-Contact;I:FM136.NET domain:Created:19950817050000000 domain:Updated:19970502190154000 %ok %ok FM136.NET contact:Class-Name:contact contact:Auth-Area:net contact:ID:FM136.net contact:Handle:FM136 contact:First-Name:Fitzgerald contact:Last-Name:Michael contact:Name:Michael, Fitzgerald contact:Type:I contact:Email:mfitzgerald@DIMENSION.NET contact:Phone:703-709-7699 contact:Org-Name:Dimension Enterprises contact:Street-Address:1039 Sterling Rd, Suite 201 contact:City:Herndon contact:State:VA contact:Postal-Code:22170 contact:Country-Code:US contact:Created:19950817123945000 contact:Updated:19960221135609000 %ok @@@@@@@@@@@@@@@@@@@@@ Jim Fleming Unir Corporation - http://www.unir.com End-2-End: VPC(Java)---C+@---<IPv8>---C+@---(Java)VPC http://www.ntia.doc.gov/ntiahome/domainname/130dftmail/unir.txt http://www.ddj.com/index/author/idx10133.htm

3 5

Re: NSI Bulletin 098-010 | Update on Whois
by Leo Bicknell 10 Sep '98

10 Sep '98

My rwhois script has been fixed and updated. The connection problem seems to be that one of the rwhois servers the Internic runs is down, but still in DNS. I have hard coded the working server for now. I added a link on the "host" output information to search for all the domains for which the host is a namesever. This is similar to the whois "SERVER <handle>" that so many people like. I have received several queries about this to know if you can get more than 256 hosts from rwhois (the limit with whois). The answer is, maybe. The internic servers have an upper limit of 50 objects on rwhois querys. You can, however run your own rwhois server and set the limit arbitrarily high, allowing someone to do this. I am attempting to setup a server to do this now, but I can't for the life of me figure out how to "mirror" the data from the Internic, or even what I need to mirror to make it happen. If someone can help me set up such a thing I'll offer (on a test basis anyway) the ability to do queries much larger than 256 hosts. The script can be seen in action at: http://www.ufp.org/~bicknell/rwhois.cgi You can download it at: http://www.ufp.org/~bicknell/rwhois.txt And, as before if you run it from the command line it produces output much like "whois". -- Leo Bicknell - bicknell(a)dimension.net Network Engineer (CCIE #3440) - Dimension Enterprises 1-703-709-7500, fax, 1-703-709-7699

1 0

RE: Inter-CO signaling?
by Chad Skidmore 10 Sep '98

10 Sep '98

RBS is often times run on on AMI D4 circuits. Also, RBS is generally not run on inter-office trunks. At least that was the case a few months ago when I worked on a DMS-500. --------------------------------------------- Chad Skidmore Director of Network Engineering Northwest Nexus, Inc. http://www.nwnexus.com 1-888-NWNEXUS > -----Original Message----- > From: Robert E. Seastrom [mailto:rs@bifrost.seastrom.com] > Sent: Thursday, September 10, 1998 11:08 AM > To: cskidmor(a)nwnexus.net > Cc: goemon(a)sasami.anime.net > Subject: Re: Inter-CO signaling? > > > > of course, robbed bit signalling and AMI have nothing to do > with one another... > > > From: Chad Skidmore <cskidmor(a)nwnexus.net> > Cc: "'nanog(a)merit.edu'" <nanog(a)merit.edu> > Date: Thu, 10 Sep 1998 10:47:04 -0700 > MIME-Version: 1.0 > X-Mailer: Internet Mail Service (5.5.1960.3) > Content-Type: text/plain > Sender: owner-nanog(a)merit.edu > > Try making an ISDN call from one to the other at a 64k > rate. If you can > get through then there are at least some trunks setup as > ESF B8ZS clear > channel. I suspect that they are probably some older > trunks running AMI > D4. > > The other thing you could do is just ask them. > > --------------------------------------------- > Chad Skidmore > Director of Network Engineering > Northwest Nexus, Inc. > http://www.nwnexus.com > 1-888-NWNEXUS > > > > -----Original Message----- > > From: Dan Hollis [mailto:goemon@sasami.anime.net] > > Sent: Wednesday, September 09, 1998 5:22 PM > > To: 'nanog(a)merit.edu' > > Subject: Inter-CO signaling? > > > > > > Is there any way to determine if inter-telco links are using > > robbed bit > > signaling? Eg from GTE to USwest in the Seattle area. > > > > I suspect RBS may be causing lots of v.90 problems in > GTE country in > > Seattle. But id like to know how to confirm this. And just > > maybe, get it > > fixed. > > > > -Dan > > > >

5 6

RE: Inter-CO signaling?
by Chad Skidmore 10 Sep '98

10 Sep '98

I you are a customer of theirs it can be hard to get them to make changes. In my past life as a CLEC employee we even had some trouble getting them to correct things that we thought were problems on inter-office trunks. You may be able to escalate your troubles up to a tech that understands and cares. If you still don't get any response I'd talk to Roger at the WUTC. ---------------------------------------------- Chad Skidmore Director of Network Engineering Northwest Nexus, Inc. http://www.nwnexus.com 888-NWNEXUS > -----Original Message----- > From: Dan Hollis [mailto:goemon@sasami.anime.net] > Sent: Thursday, September 10, 1998 11:39 AM > To: Chad Skidmore > Cc: 'Robert E. Seastrom'; 'nanog(a)merit.edu' > Subject: RE: Inter-CO signaling? > > > On Thu, 10 Sep 1998, Chad Skidmore wrote: > > RBS is often times run on on AMI D4 circuits. Also, RBS is > generally > > not run on inter-office trunks. > > I have a feeling RBS is being used for inter-telco trunks (eg > USwest to > GTE). And its killing v.90 in Seattle. > > If a telco is running RBS on a trunk, any way to get them to > remove it? > > Also, any way to get the telco to tweak digital pads? Eg what magic > incantation do you say to the telco. > > -Dan >

1 0

Re: NSI Bulletin 098-010 | Update on Whois
by Sean Donelan 10 Sep '98

10 Sep '98

The WHOIS database has a long history traced back through several generations of managers to the early days of ARPA, DCA, IANA, SRI. I don't think anyone really has clear title to claim 'ownership' of the data. The database has been transfered several times already from one contractor to the next. SRI transfered it to GSI, and then to NSI. The database has been split several times. Such as when .MIL was split off, when .GOV was split off, when ARIN was split off. Usually these splits have ended up screwing up the NICHANDLES of those ancient individuals on the net. Giving NSI the benefit of the doubt, I assume NSI will transfer the database to whatever the successor organization (or organizations) is, as NSI has previous done with the .GOV data, and NET-BLK data. As far as the only supported interfaces for retrieving the data from NSI being the WHOIS port and the anonymous FTP zone files, if you look at it in the best light, I would refer back to the cooperative agreement with NSF. As much as I dislike SPAM, I do not think concern about the use of the data for marketing purposes is a good justification for NSI to unilaterally block individuals from accessing the WHOIS data. I would be interested in knowing why a couple of sites have found the need to query the WHOIS data so heavily. And once we knew their need, perhaps finding a better solution. But the NSF cooperative agreement didn't really say much about finding new ways to efficiently distribute the WHOIS database. But, as folks around the beltway like to say, its not NSI's bailiwick. If NSI is looking for a nice off-site storage facility to store a backup of the database in case something happens on the east coast (see http://nic.ddn.mil/DNS/root-server.html) I know of a nice centrally located facility :-) -- Sean Donelan, Data Research Associates, Inc, St. Louis, MO Affiliation given for identification not representation

7 8

Ok, perhaps the InterNIC's Chuck Gomes was lying?
by Steven J. Sobol 10 Sep '98

10 Sep '98

This surprises me. And pisses me off. Dean Robb pointed to some SEC filings, a prospectus from NSI which he gave an incorrect URL for -- the correct url is http://www.sec.gov/Archives/edgar/data/1030341/0000950133-97-003381.txt which shows that Network Solutions, Inc. (the InterNIC people, for those of you reading on Freemail who are not familiar with the name) plans to build a database of customer information that "makes proper use of the customer data that the Company collects." Doesn't specifically list selling e-mail addresses, but... Now, Mr. Gomes assures me that the InterNIC has no plans to spam. Maybe this is true. However I will state right here and now that *none* of my e-mail addresses had better be sold to anyone. NSI does not own them. Furthermore, they are used for BUSINESS PURPOSES and I can't just start blithely filtering e-mail. Mr. Gomes has also told me that SAIC is just a shareholder, but this is part of the prospectus: CLASS A COMMON STOCK Of the 3,300,000 shares of Class A Common Stock offered hereby, 2,800,000 are being sold by Network Solutions, Inc. ("NSI" or the "Company") and 500,000 shares are being sold by the Company's parent, Science Applications International Corporation, a Delaware corporation ("SAIC" or the "Selling Stockholder"). (note that SAIC is listed as the PARENT COMPANY of NSI). Now, the second URL. http://www.sec.gov/Archives/edgar/data/1030341/0000950133-98-001406.txt As mentioned by Dean, a bunch of people on NSI's board are also on SAIC's board. I'm giving Chuck the benefit of the doubt, for now, since he has been helpful and (apparently) straightforward, but the prospectus is less than a year old (it's dated 9/26/97), so the information should be current. The proxy statement is dated 4/14/98. It'd be great if Chuck would be willing to subscribe to NANOG and talk to us here. (This is being CC'd to him) -- "You know, I've decided lizards aren't too smart." --Me, to "Junior", one of my iguanas

2 1

Atlanta Meeting in November
by Susan Rebecca Harris 10 Sep '98

10 Sep '98

Greetings everyone - registration is now open for the fall NANOG conference, to be held November 8-10 in Atlanta: http://www.nanog.org/mtg-9811/update.html Hope you can join us! If you have questions about registering, please contact Pam Ciesla (pam(a)merit.edu or 734.936.0172.) The NANOG registration fee has now been set at $250, up from $150. The increase will allow Merit to cover catering and A/V costs, as outlined by Bill Norton this past June. You'll find details of the new funding model at: http://www.nanog.org/budget.html Many thanks to Netrail for hosting the meeting, and to the organizations co-sponsoring meals and events at the conference. --susan harris

1 0

saudi internet / international connections
by Peter Merdian 10 Sep '98

10 Sep '98

Hello, if this is not the right list, please tell me so. Otherwise: Saudi Arabia will connect the whole country through one party (KACST), in the beginning to UUNET in the US. I'd like to have an advice, whether it is better to connect to - an internat. exchange / NAP/MAE (and have the transit/uplike and peering there) - or directly to a large ISP like UUNET/MCI/Sprint? Out of redundancy (not financial) reasons there will be a link to Europe. So the same questions like above applies and additional the question whether to connect to TEN-34 or EBONE - which are supposed to be a/the European backbone. How are the chances today to have some peerings and not going all the traffic through an transit/uplink ISP? On a sidenote: it is 50% cheaper to dump the traffic in the US than in Europe. Is it (for European ISPs) cheaper to have raw bandwidth to US (from Europe) plus uplink-ISP fee than connecting to e.g. MAE-Frankfurt ?! There seems to be some way still to go for a true worldwide Internet. With regards, Peter Merdian -- currently at KACST, Saudi Arabia

2 1