- Test - lists.nanog.org

Response to NOI from NTIA on DNA
by Jay R. Ashworth 10 Jul '97

10 Jul '97

... and if the VC gets the VP who's here on the QT we'll all get put on KP. :-) Follows a fairly long message containing my comments and opinion subject to a Notice of Inquiry issued by the US Government's NTIA (which means something, but I forget what.) The rough idea as I understand the NOI is "you Internet people got important, and you seem to be having trouble sorting out your domain name registration management stuff. Do we need to step in and help you?" (read: tell you how to do it) I'd appreciate any comments y'all care to make on this, either technically, stylistically... or simply pointing out stuff that my spell checker missed. ;-) My intent is to hit somewhere about halfway between formal corporate comments and much of the self-interested, poorly thought out garbage the website show that they've been getting. http://www.ntia.doc.gov/ntiahome/domainname/domainname.htm (which URL worries me all by itself, but...) Cheers, -- jra -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Follow my comments on the Notice of Inquiry concerning Domain Name Service management. I speak as a commercial consultant on internetworking and computer systems design with 15 years experience. I've used the "call and response" format customary to Internet email; hopefully, my responses will be comprehensible, rather than compost. > Various private sector groups have proposed systems for allocating and > managing generic top level domains (gTLDs). The Government is studying Actually, the initialism "GTLD" is customarily understood to mean "Global Top Level Domain". > A. Appropriate Principles > > The Government seeks comment on the principles by which it should > evaluate proposals for the registration and administration of Internet > domain names. Are the following principles appropriate? Are they > complete? If not, how should they be revised? How might such > principles best be fostered? > > a. Competition in and expansion of the domain name registration system > should be encouraged. Conflicting domains, systems, and registries > should not be permitted to jeopardize the interoperation of the > Internet, however. The addressing scheme should not prevent any user > from connecting to any other site. > > b. The private sector, with input from governments, should develop > stable, consensus-based self-governing mechanisms for domain name > registration and management that adequately defines responsibilities > and maintains accountability. > > c. These self-governance mechanisms should recognize the inherently > global nature of the Internet and be able to evolve as necessary over > time. > > d. The overall framework for accommodating competition should be open, > robust, efficient, and fair. These first four points are platitudes... but they're well thought out platitudes. :-) > e. The overall policy framework as well as name allocation and > management mechanisms should promote prompt, fair, and efficient > resolution of conflicts, including conflicts over proprietary rights. This is well phrased, but will be quite difficult to manage in practice. The precise reasons why, I'll take up shortly. > f. A framework should be adopted as quickly as prudent consideration > of these issues permits. In light of the current travails with NSF contractor Network Solutions, yeah, this is a good idea, too. > B. General/Organizational Framework Issues > > 1. What are the advantages and disadvantages of current domain name > registration systems? >From an operational standpoint, the major disadvantage to the current registration system is that Network Solutions appears, based on reports from a statistically significant number of it's customers, incapable of providing reasonable customer service, from a standpoint of response time if no other. If a request is at all out of the ordinary, the systems in place at NetSol are guaranteed to both mishandle it and delay it. Even normal requests are often troublesome. For example, NSI apparently has an undocumented policy of allowing the "Host Registration" necessary to use a host as a DNS server to be done only once for any given IP address. I'm involved in a situation right now where administrative control of a network is about to change hands, and I attempted to register two hosts in that domain to use as servers. The two registration templates templates were sent out within 60 seconds of one another, the "automated acknowledgement" messages came in 2 and 9 hours later respectively, and 3 _days_ later, _one_ of the registrations was acknowledged. The other registration vanished into limbo, and 60 minutes of toll phone calls a week later were necessary to discover that the registration was bounced "because another host is already registered with that address". Inasmuch as the very DNS system these registrations were intended to support allows multiple names for one address, this is puzzling, but the lack of response, the lack of documentation, and my total inability to successfully contact anyone in authority to discuss the matter are unconscionable. It _has_ to be possible to provide better customer service than this... with a $50M annual revenue stream. >From a _structural_ standpoint, the current Domain Registration system is deficient primarily in it's centralization. Many other components of the Internet have evolved over the course of the last 20 years, but DNS and registration administration is just now reaching it's adolescence. The other major, and probably insoluble, problem is that the DNS system uses names as addresses. The problem is that names can change, and addresses usually shouldn't. The quintessential example of this is professional service companies like law firms. When Trenam, Simmons, Kemker, Scharf, Barkin, Frye and O'Neill loses a partner and becomes Trenam, Kemker, Scharf, Barkin, Frye, O'Neill and Mullis, what do you do with all that stationary... and more to the point, all those bookmarks and web index engine entries, that say "trensim.com" (or "tsks.com", or whatever)? But it's _way_ too late to do anything about this now, and I'm not sure there ever was a time when it wasn't. > 2. How might current domain name systems be improved? Caution is needed here; what's at odds is primarily the registration systems that underlie DNS, not the technology itself. It's difficult to answer this question without getting "personal" about NetSol; the primary improvement I can see at the moment would be for NSI to start earning the incredible amount of money they unilaterally decided to charge for domain registration service. The other answer to this question is global, and the primary target of this NOI; I'll return to it after building some more ground work. > 3. By what entity, entities, or types of entities should current > domain name systems be administered? What should the makeup of such an > entity be? There have been half a dozen proposals made for a restructuring of the DNS registration services infrastructure. Of all the approaches I've investigated, I believe that the Denninger/Postel Internet Draft on the topic is the best thought-out, and most comprehensive. Extreme care is necessary here: one of the reasons that the Internet has successfully scaled to the degree that it has in the short amount of time it took is that the underlying foundations of the protocol designs and their implementations were subject almost entirely to engineering discipline; commercial and (say it softly) political concerns were ignored. Make no mistake, the expansion of the DNS registration infrastructure which everyone agrees is necessary must take commercial concerns into account... but it _MUST_ be designed by engineers; it's an engineering issue. We don't allow politicians to design interstate highways. > 4. Are there decision-making processes that can serve as models for > deciding on domain name registration systems (e.g., network numbering > plan, standard-setting processes, spectrum allocation)? Are there > private/public sector administered models or regimes that can be used > for domain name registration (e.g., network numbering plan, standard > setting processes, or spectrum allocation processes)? These issues are covered in the Denninger/Postel draft, but I'll note that while the primary concerns are infrastructural, and thus engineering, the main secondary, operational, concern is that of validation of registration entities, providing for a common set of clearly enumerated policies (for things like trademark disputes) which all registration entities must agree upon(/have imposed on them). > What is the > proper role of national or international governmental/non-governmental > organizations, if any, in national and international domain name > registration systems? >From an operational standpoint, there _must_ be some centralized agency with responsibility for the "ownership" of the root of the DNS namespace (commonly, but incorrectly, referred to as "."). However, this agency's sole duty should be to delegate it's authority to TLD registries and arbitrate disputes. The design of the system and the charter of this board should be such as to make it structurally immune to litigation about issues like, for example, trademarks. It must have both the authority and the resources to reassign or temporarily support any domain whose registrar become unable to continue it's services. Two points are important here: 1) Registry services and DNS service provision are related but need not be combined: it's possible to envision an environment in which registries contract out the actual provision of DNS root services to a technically competent third party, thus isolating customers from business problems at the registry entity, and 2) Regardless of the legalities, domain names are being viewed as property by their holders, and substantial investments are being made in them, primarily in publicity, but also in customer mindshare. The nature of the net is such that it depends on this behavior, and therefore it must be taken into account when prioritizing such items as continuance of service. > 5. Should generic top level domains (gTLDs), (e.g., .com), be retired > from circulation? I think not, for the reasons enumerated above. Many companies are actually _named after_ their domain names; and while "no law guarantees that anyone will be able to continue making his living in a certain manner" (Judge Learned Hand), neither are flag days looked upon kindly; justification for such things must be reached by consensus. > Should geographic or country codes (e.g., .US) be > required? See above; ie: no. > If so, what should happen to the .com registry? Are gTLD > management issues separable from questions about International > Standards Organization (ISO) country code domains? I don't see any good reason to need to separate the topics; the parallel structures don't seem to be what is causing the problem. > 6. Are there any technological solutions to current domain name > registration issues? Are there any issues concerning the relationship > of registrars and gTLDs with root servers? The current issues appear to be architectural and commercial, rather than technical, and therefore require architectural solutions. > 7. How can we ensure the scalability of the domain name system name > and address spaces as well as ensure that root servers continue to > interoperate and coordinate? These are mostly technical questions, and I don't claim to be an expert on the topic, but if I don't see the names Vixie, Halley, Margolin, and Liu on any paper asserting to answer this question authoritatively, I'll assume it doesn't know what it's talking about. That is: there are experts on these topics, and anyone in authority who flouts them, or worse, ignores them, does so at the peril of the entire Internet. (Note to readers: there are other DNS experts, obviously; I simply picked the top 4 I see on the mailing list as examples.) > 8. How should the transition to any new systems be accomplished? This is already taking place. There are root nameservers which are _not_ authoritative for .com and the other domains currently run by NSI; this experiment seems to be working. There are also currently operational root servers for view of the namespace which include alternate TLDs, these include alternic.nic. In short, as long as the current operators of the DNS roots (which primarily means NSI) help rather than hinder, a transition will be a Small Matter of Administration. > C. Creation of New gTLDs > > 10. Are there technical, practical, and/or policy considerations that > constrain the total number of different gTLDs that can be created? Mostly, the size of the TLD tag. Traditionally, these have been 2 or 3 characters; the D/P draft suggests a maximum of 4 or 5, which seems sand, but does impose an absolute limit. Also, this limit is smaller than it might seem it ought to be: all components of domain names _must_ be pronounceable. This is more a social limitation than a technical one (indeed, the software doesn't care), but it's a requirement nonetheless. > 11. Should additional gTLDs be created? Let's be careful here: the same namespace issues apply to TLD's that apply to Usenet newsgroups: creating new ones without extensive discussion and justification ought to be _expensive_ and _time consuming_. Much too little concern is given to namespace control... which isn't surprising; it's an architectural issue, and most people aren't architects. But it's nonetheless crucial to the ongoing simplification and "consumerizing" of the net of the net. Case in point example: ".firm". Is there really anyone who's _thinking_ about this, who doesn't know which company will register "ibm.firm" the second it hits the table? They've probably got an employee whose job is nothing else, by now. That is to say, "horizontal" segmentation of the namespace will not work; the problems are identical to those in the botched release of the 888 toll free NPA. Who owns 1 888 FLOWERS? And ".nom" is simply stupid. However, in the grand scheme of things, yes. > 12. Are there technical, business, and/or policy issues about > guaranteeing the scalability of the name space associated with > increasing the number of gTLDs? Well, it's likely that as long as the scaling doesn't go too fast, the technology and policy issues involved can keep up with it. Ensuring this is probably the job of whatever group gets appointed to own the root of the namespace. And I'll say this again. This is an architectural function. Would _you_ want to live in a house designed by a politician? > 13. Are gTLD management issues separable from questions about ISO > country code domains? No; ISO3166 registries already exist, and by their nature, probably should have their policy making continue unimpeded by any except technical considerations. The only control that appears necessary is the "we're the new government, delegate to us now" sort of incident... and this is large enough that it doesn't really matter _who_ has the responsibility... professional diplomats will be the implementors. > D. Policies for Registries > > 15. Should a gTLD registrar have exclusive control over a particular > gTLD? Are there any technical limitations on using shared registries > for some or all gTLDs? Can exclusive and non-exclusive gTLDs coexist? I should think that it would be difficult to have more than one registrar for a TLD. The difficulty can be better illustrated by observing that there are three functions performed by registrars: 1) Policy and administration, 2) registration operations, and 3) nameserver operation. The first is the largest problem, and for technical reasons, the current implementation of DNS makes division of number 3 difficult as well. > 16. Should there be threshold requirements for domain name registrars, > and what responsibilities should such registrars have? Who will > determine these and how? Yes, there should. There is some merit to the idea, possibly original to me, that these threshold requirements should be a contractual issue between the registry and its clients (with the exception of grandfathered TLD's, of course). Since these requirements are to protect the clients, the only reason I can see for externally-imposed requirements are in the event that more than one entity applies to host the same TLD name simultaneously. Arbitrating these types of disputes would be another job of the root operators. > 17. Are there technical limitations on the possible number of domain > name registrars? Only the available number of registerable TLDs. > 18. Are there technical, business and/or policy issues about the name > space raised by increasing the number of domain name registrars? Hmmm... technically, yes. It becomes necessary to uncouple the root nameservers from the TLD nameservers. Business? Making sure that customers see a _reasonably_ coherent view of the TLD namespace, from a registration policies standpoint. > 19. Should there be a limit on the number of different gTLDs a given > registrar can administer? Does this depend on whether the registrar > has exclusive or non-exclusive rights to the gTLD? I believe that this is a question of workload and the ability to handle it. I _do_ think that a raw numerical limit wouldn't work the way it was intended, as there are groups of TLD's that comprise a "concept", which probably ought to be administered together... like, for example, .am, .fm, .tv, .news and .mag. These also comprise a good example of TLDs which ought to have special policies; I, for example, would allow in the first three categories only customers who could document an FCC broadcasting license for the appropriate domain. > 20. Are there any other issues that should be addressed in this area? Probably, but I'm pretty certain that the D/P draft addresses them. > E. Trademark Issues Oh, God. > 21. What trademark rights (e.g., registered trademarks, common law > trademarks, geographic indications, etc.), if any, should be protected > on the Internet vis-a-vis domain names? This is probably the single biggest problem with the current .com domain. As long as the policy is stable, and well documented, and _doesn't change on a whim_, I'm not sure it matters. The market will fix any inequities here. The reason this is really a problem, though--as is a surprise to no one--is one of jurisdiction. There are hundreds of geographical jurisdictions for trademark control, and usually, the geographical separation involved is enough. "Smith's Plumbing" in Alaska probably cares very little about competition from "Smith's Plumbing" in Arkansas. And then along came the net. There's no perfectly satisfactory solution... But disabling a domain name's service that a customer has both paid for and advertised extensively, without warning, or right of contest -- NetSol's policy -- is simply _not_ acceptable. > 22. Should some process of preliminary review of an application for > registration of a domain name be required, before allocation, to > determine if it conflicts with a trademark, a trade name, a geographic > indication, etc.? Nothing would ever get registered. _Everything_ conflicts with something, somewhere on the globe. In the course of ordinary business, these factors are the responsibility of the business, I see no reason why they shouldn't stay there. > If so, what standards should be used? Who should > conduct the preliminary review? If a conflict is found, what should be > done, e.g., domain name applicant and/or trademark owner notified of > the conflict? Automatic referral to dispute settlement? Anything except the sudden "On Hold" will be fine, thanks. > 23. Aside from a preliminary review process, how should trademark > rights be protected on the Internet vis-a-vis domain names? What > entity(ies), if any, should resolve disputes? Are national courts the > only appropriate forum for such disputes? Specifically, is there a > role for national/international governmental/nongovernmental > organizations? Alas, (I say alas because countries have a disturbingly long history of not being able to agree on these topics), I don't think anything except an international body of some kind will have jurisdiction. > 24. How can conflicts over trademarks best be prevented? What > information resources (e.g. databases of registered domain names, > registered trademarks, trade names) could help reduce potential > conflicts? If there should be a database(s), who should create the > database(s)? How should such a database(s) be used? This is actually a question of intellectual property rights management, not one of technology per se, nor the Internet, per se. Check with Carl Oppedahl; he makes a living on this stuff. :-) > 25. Should domain name applicants be required to demonstrate that they > have a basis for requesting a particular domain name? If so, what > information should be supplied? Who should evaluate the information? > On the basis of what criteria? This is a difficult question to answer. It might help, but I feel that the net is a hotbed of opportunity for entrepreneurialism, and entrepreneurs often do the legal paperwork last. Besides, this would impose one more load on potential registrars. No, it's probably a good idea, but I'm not sure it's feasible. > 26. How would the number of different gTLDs and the number of > registrars affect the number and cost of resolving trademark disputes? I think the root-ops would have to make the appropriate arbitration policies and require TLD registries to adopt them as part of their contracts. > 27. Where there are valid, but conflicting trademark rights for a > single domain name, are there any technological solutions? Nope. Modify the name in some fashion. A client is an Allied Van Lines moving agent. When Allied went to request a domain name, {allied,avl}.{com,net} were all taken. On bad advice, they registered alliedvan.net, which they've since modified to the more expected alliedvan.com. It isn't great, but the other registrants had good trademark claims to the other names, as well, so Allied Did The Right Thing, and solved it's problems by clever _use_ of the technology, rather than in court. Note that that's not a "technological fix"... it's simply an intelligent _use_ of technology that already exists. Another example: that client is on the net as well. When they went to apply, they discovered that "blocker.com", the appropriate domain name for this 99 year old company, was already taken... by some "name registry company" in Canada for an unknown, and possibly non-existent client. So, taking advantage--again--of a little common sense, they registered "blocker100.com", "in recognition of their upcoming century anniversary". > 28. Are there any other issues that should be addressed in this area? Dozens, but they haven't all come up yet. Copies of the Denninger draft are available from www.alternic.net, the Postel inet-draft is in the usual places. And in closing, allow me to compliment NTIA for soliciting, and indeed _allowing_ the submission of comments electronically; this much improves the chances you'll get what you're looking for. Hopefully, my comments will prove useful, informative... and not too derogatory to NSI. :-) Cheers, -- jra -- Jay R. Ashworth High Technology Systems Consulting Ashworth Designer Linux: Where Do You Want To Fly Today? & Associates ka1fjx/4 "...short of hiring the Unabomber, how can I +1 813 790 7592 jra(a)baylink.com get back at them?" --Andy Cramer NIC: jra3 -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- -- Jay R. Ashworth jra(a)baylink.com Member of the Technical Staff Unsolicited Commercial Emailers Sued The Suncoast Freenet "People propose, science studies, technology Tampa Bay, Florida conforms." -- Dr. Don Norman +1 813 790 7592 -- Jay R. Ashworth jra(a)baylink.com Member of the Technical Staff Unsolicited Commercial Emailers Sued The Suncoast Freenet "People propose, science studies, technology Tampa Bay, Florida conforms." -- Dr. Don Norman +1 813 790 7592

2 1

Re: Keynote/Boardwatch Results
by Jack Rickard 10 Jul '97

10 Jul '97

Avi: "Cheating" is of course encouraged. This isn't an academic test at your local university. We're all out of school now. If you can figure out a way to beat the game, you have ipso facto figured out a way to make the web look faster to end users. As it appears to be, so it is. As Martha Stewart says, that would be a "good thing." It would be a good thing for you and your product line. It would be a good thing for your web site customers. It would be a good thing for end-users viewing such web sites. If the net effect of all this is that all the smart people on the net get mad at me and go figure out a way to make the web look faster, it will be well and good enough for me as well. I had previously agreed NOT to publish IP numbers of the Keynote host machines. Keynote does make this information available on their web site, so I myself was a little bemused by the request, but I did agree to honor it. In any event, someone else has already posted the locations and networks (which we DID publish), along with the IP numbers, here on the list. So you should have them. If mirroring/caching moves the numbers definitively, it then establishes a "real" value to such a technique, and it can be offered to customers at a higher price with some actual data comparing how they will "look" to the world using both the less expensive simple hosting as compared to the more expensive geographic mirroring technique. I personally think this would move the numbers more than anything else that could be done, but that's what looks LOGICAL to me, not anything I know or have tested. I am rather convinced that moving a single web site to another location, putting it on larger iron, or using different OS will have very minor impact on the numbers. My own personal theory is a little vaguely formed. But I think the heart of the performance problems lie in a visceral connundrum of the network. It is and should be one network. The perceptual disjuncture I already detect, even among NANOG members, between straight "routes" as viewed by traceroute and ping, and the way data actually moves on the network (at least vis a vis the mental model or "theory" I have of it) is a somewhat shocking part of the problem. I was actually unaware that many of the network engineers actually viewed the world in this way until this exercise. I was even a bit flip about not dealing with ping/traceroute at any level of comparison. Perhaps an article on this is in order. But I think most of it has to do with interconnects between networks, and architectural decisions accummulated over the years based on concepts of what should be "fair" with regards to the insoluble, but ever moronic "settlements" concept and who gets value from whom based on what. If decisions had been more based on optimizing data flows, and less on whose packets transit who's backbones and why, performance would have been improved. I don't know how much, but certainly some. When the main thing on the minds of CEOs of networks is preventing anyone from having a "free ride" (ala SIdgemore's theory of optimizing Internet performance by it being owned totally by UUNET), or the relatively mindless routing algorithm of moving a packet to the destination network at the earliest opportunity to make sure "I" am not paying for it's transit, if it goes to "your" location, I suspect performance suffers. My sense is larger numbers of smaller networks, interconnected at ever more granular locations, would be a good thing. This will get me in big caca with the "hop counting" mind set, and of course at about 254 hops a minor little problem arises, but I think so nonetheless. Very small ISPs know this viscerally. They all want to multihome to multiple backbones, and have done some work to interconnect among themselves locally. Savvis actually has a very interesting concept though it upsets everybody. It kind of upsets me because it makes my head hurt, literally. They've carried it almost to another level of abstraction. If you ponder it long beyond the obvious, it has some interesting economic consequences. Checkbook NAPS lead to an inverted power structure where the further away you move from centralized networks such as internetMCI and Sprint, by blending layers after the fashion of a winemaker, the better your product becomes and the better apparent connectivity your customers have. The head hurting part is that if you extend this infinitely we would all wind up dancing to the tune of a single dialup AOL customer in Casper Wyoming somewhere in the end. But there is a huge clue in here somewhere. In all cases, Savvis numbers were better than the either UUNET, Sprint, or internetMCI numbers individually. Would it then be true, that if there were three Savvis's each aggregating three networks with a private NAP matrix, and I developed a private NAP matrix using the three Savvis level meshes, would my performance be yet better again? And what if Savvis opened the gates and allowed UUNET end users to connect to Sprint IP Services web sites transiting via Savvis network? More vaguely, if you have four barrels of wine with one a bit acidic, one a bit sweet, one a bit oakey, and one a bit tannic, and you blend them all together, it would appear apparent that you would have a least common denominator wine that is acidic, sweet, oakey, and tannic. You don't. You get a fifth wine that is infinitely superior than the sum of the parts or any one component barrel. It is an entirely "new thing." This is sufficiently true that it is in almost all cases how wines are made now. Are networks like wine? Jack Rickard ============================================================== Jack Rickard Boardwatch Magazine Editor 8500 West Bowles Ave jack.rickard(a)boardwatch.com Littleton, CO 80123 (303)973-6038 voice (303)973-3731 fax http://www.boardwatch.com ============================================================== ---------- > From: Avi Freedman <freedman(a)netaxs.com> > To: Jack Rickard <jack.rickard(a)boardwatch.com> > Cc: mohney(a)access.digex.net; nanog(a)merit.edu; GeneShklar(a)keynote.com > Subject: Re: Keynote/Boardwatch Results > Date: Tuesday, July 08, 1997 9:40 PM > > > It would appear that everyone is pretty smugly satisfied by concensus that > > the performance series we ran actually measures server performance and that > > since all ISPs run weeny home servers, this was not "really" a test, flawed > > methodology, etc. I corresponded with Doug the Hump at Digex about this. > > I've liked this guy since I first met him largely because he's funny and > > doesn't take himself too seriously. He's got a yen for black helicopters > > that still has me in stitches. > > Humph. Doug the Hump indeed. Well, Alex @ our shop wants an Apache > Helicopter for our NOC as well. I'm laughing because I'm not sure anyone's > called him that before... > > Anyway, I'm thinking of putting www.netaxs.net on one of our core routers :) > > Think that'd help? > > Actually, what we'd do is make it a loopback interface on all of our > core routers and thus you'd hit whatever the closest router to the > querying machine is, bypassing much of the network. > > Hmm. Have to try that one out. > > Anyway, I have to take a look at some of the test sites (saw some of > them listed in the new Directory) and see if I can figure out some > of the topology of the testing. > > Jack - could you put up IPs or whatnot of the sources (the test sites) > for people to "tune" and test to? > > > jack.rickard(a)boardwatch.com Littleton, CO 80123 > > Avi

5 5

ATM vs. DS3
by Josh Beck 10 Jul '97

10 Jul '97

Hello, I just thought of something. We are in the process of purchasing a 4 Mb CIR from another backbone. Now, we have the choice of ATM or standard T3 delivery (over a DS3 either way). Now, if we get ATM, that 4 Mb CIR turns into: [ (53-5)/53 ] * 4 Mb/s = 48/53 * 4 Mb/s = 3.62 Mb/s Because an ATM cell is 53 bytes, but 5 bytes of that is header. Thus we pay for 4 Mb/s and get 3.62 Mb/s. Now, the question is, what is the overhead of the serial (ppp?) protocol running over the T3? Since it is point to point, and not addressed, it should be less, or is this not the case? Any input would be appreciated! Josh Beck jbeck(a)connectnet.com ------------------------------------------------------------------------ CONNECTNet INS, Inc. Phone: (619)450-0254 Fax: (619)450-3216 6370 Lusk Blvd., Suite F-208 San Diego, CA 92121 ------------------------------------------------------------------------

2 1

Re: Suspension of reissuing deleted domain names
by Francois Beauregard 10 Jul '97

10 Jul '97

And what do you do when the domain was not paid.... you block a potentially ligit user from taking it over ??? At 22:06 97-07-09 -0400, Daniel Senie wrote: >With regards the issue of prospectors trying to grab deleted domains, >perhaps the InterNIC should use a similar policy to what the FCC uses >for radio callsigns (amateur radio, at least): Callsigns which are >abandoned due to change of callsign, death, or expiration of the license >are unavailable for re-issue for two years. This allows for the previous >owner (if not dead) to reclaim the callsign (or a relative of the >deceased to do so). It also helps avoid identity confusion. > >Seems like the same rules would make sense in the domain name context, >and would avoid problems where billing errors or other problems dealing >with the Internic could result in loss of a domain name. Sincerely ---------------------------------- Francois Beauregard FBorg(a)fbli.com FBLI.COM We love feedbacks and live by them Sales : Sales(a)fbli.com 1 (888) FBLI.COM Tech support : Support(a)fbli.com 1 (514) 349-0455 Internet Web : http://www.fbli.com ICQ : 1907537 Montreal, Canada ----------------------------------

1 0

Re: Internet Backbone Index
by Sean Donelan 10 Jul '97

10 Jul '97

>SAVVIS. Fortunately Our target markets are not just libraries and other >information providers, it's EVERYONE that needs a T1 and above connection >to the Internet. How many cities are you in Sean, where are DRA's POPs for >customers to access? How much bandwidth does DRA have to get these >customer to other network? Let's compare bandwidth shall we. DRA tries to do one thing well, rather than a lot of things not so well. Like all things, the correct answer is "It depends." If you are interested in the DRA network geography, a pretty picture of the DRA North American POPs is at <http://dranet.dra.com/dranet.html>. BTW, all those locations are up, operational, and have DRA owned and operated equipment in place. DRA's international offices are located in Montreal Canada, Paris France, and Melbourne Australia. The primary NOC is in St. Louis, MO and backup NOC operates out of Monterey, CA. Since DRA also has a lot of private connections (in the old NSFNET AUP days these were called 'backdoor' connections) calculating total bandwidth is a bit confusing. If you follow the AGIS bandwidth counting method, DRA has about 268 Mbps of possible bandwidth to other networks through public and private inter-connects, although DRA doesn't have any single link faster than 45 Mbps. However, bandwidth is rarely the problem. I've found a well engineered 56Kbps connection outperformed a DS3 port into a poorly engineered network. >When 80 to 90 percent of the Internet traffic is to MCI, SPRINT and UUNET >then our model is the right way to build this, not to try and see how many >peering agreements one can get. Except the model falls apart if 50% percent of the your Internet traffic isn't just to MCI, Sprint and UUNET, as in DRA's case. If you want to get anywhere off the commercial beaten path, like many of our customers do, things quickly get very bad if you stick with just those three providers. Most of our backdoor connections exist exactly for that reason. Look at the KEYNOTE test sites. Although it might appear heavly weighted towards MCI, SPRINT, and UUNET (23 out of 35 sites), the network is never so simple. It works well as long as you are in a US city going to another US city. But looking at the Bell Canada graph, things don't work as well. Hint: If you are trying to reach a Canadian audience, don't put your web server in a US city. We've had a private 'backboor' connection to the University of Toronto for years for precisely this reason. Further, about half the Keynote test sites seemed to have alternate connections. I couldn't figure out the numbers until I started doing traceroutes, and then things started making more sense. You'll discover the best route isn't always through MCI, Sprint or UUNET. Those little exchange points can make a difference. St. Louis, MO MCI 207.230.62.16 Cybercon/Internet 1st $ traceroute 207.230.62.16 traceroute to 207.230.62.16 (207.230.62.16), 30 hops max, 38 byte packets 1 StLouis22-e3.dra.net (192.65.218.2) 200 ms 10 ms 0 ms 2 stlouix.starnet.net (198.32.132.12) 10 ms 20 ms 0 ms 3 e0-1-2.starnet2.starnet.net (199.217.255.97) 70 ms 10 ms 10 ms 4 router.cybercon.com (199.217.252.58) 10 ms 20 ms 10 ms 5 207.230.62.16 (207.230.62.16) 20 ms 10 ms 10 ms It is amusing to look at AGIS's chart in the Boardwatch directory. The graph dramatically demostrates how badly AGIS's tough-line peering policy hurt it in this test. Refusing to peer with CRL or Goodnet wouldn't have changed CRL's or Goodnet's performance very much, but it does appear to hurt AGIS's performance. Maybe Sprint and MCI might want to re-think their peering policies also. On the other hand, I've never heard of CompuServe turning down an opportunity to inter-connect their network with other networks. I'm a pragmatic person. DRA has connections to several exchange points, and has many more private 'backdoor' connections to other networks. DRA even buys (gasp) service from a few other providers, and also sells service to a few other providers. All work well, customers are happy, and DRA is profitable. Let's compare balance sheets shall we? -- Sean Donelan, Data Research Associates, Inc, St. Louis, MO Affiliation given for identification not representation

1 0

Re: UUNet 10Plus
by Sanjay Dani 10 Jul '97

10 Jul '97

>Which NAP is that? http://www.ix.digital.com >>know the customer can't push 10Mb anyway. He "pretty much" refused to >>give the 10 plus service within a NAP --where our router is colocated

1 0

Re: UUNet 10Plus
by Alex Rubenstein 09 Jul '97

09 Jul '97

Which NAP is that? At 12:20 PM 7/9/97 -0700, Sanjay Dani wrote: > >>questions, or comments would be appreciated. Once again, I understand >>that you only get 6Mbps out of a 10Mbps ethernet link, I just want to know >>why UUNet won't deliver a better handoff to their 10Plus customers and >>allow them to actually get 10Mbps. I guess they're going to try pushing > >The last time I talked to a UUnet sales rep with a clue, he >"pretty much" acknowledged that the reason their 10 Plus service >is not much more expensive than the dual T1 service is because they >know the customer can't push 10Mb anyway. He "pretty much" refused to >give the 10 plus service within a NAP --where our router is colocated >with theirs and PNIs are allowed-- over a 10baseT (better than >the netedge) or local loop DS3 handoff. > >Sales people never say things directly, hence the "pretty much" quote >above which is my reading of his excuses. > >Sanjay. > --- "Don't go with a spineless ISP; we have more backbone." Alex Rubenstein -- alex(a)nac.net -- KC2BUO -- www.nac.net net @ccess corporation, 201-983-0725 -- 201-983-0725

1 0

Re: UUNet 10Plus
by Alex Rubenstein 09 Jul '97

09 Jul '97

Also; can you get the NetEdge to do Full Duplex? That could improve your througput. At 03:17 PM 7/9/97 -0400, Deepak Jain wrote: > >My impression [which could be wrong] was that the 10Plus service was >delivered over clear channel DS3 between netedges and the actual 10Mbit/s >portion was dealt with at the media exchange. I didn't know/think they >formatted the DS3 as an ATM 10mbit/s circuit. > >Then again, I have never ordered 10Plus service. > >-Deepak. > >On Wed, 9 Jul 1997, Joe Shaw wrote: > >> Well, the ATM connection to the Net Edge box is only setup to do 10Mbps, >> so at 6Mbps on ATM, you've filled your cells. I think the whole ideaology >> behind the thing is flawed, and if they're going to sell something they >> call 10 Plus, they need to at least provide 10 of something other than >> 10Mbps in theory. I'm not happy with UUNet (it took 5 months to get the >> Net Edge box replaced), and I doubt they will have anything but the >> smallest possible presence in our network in the future. >> >> Joe Shaw - jshaw(a)insync.net >> NetAdmin - Insync Internet Services >> "Learn more, and you will never starve." - Paraphrase of Lee >> >> On Wed, 9 Jul 1997, Deepak Jain wrote: >> > >> > You could always plug an FDDI card into the back of the Netedge. >> > >> > -Deepak. >> > >> >> > --- "Don't go with a spineless ISP; we have more backbone." Alex Rubenstein -- alex(a)nac.net -- KC2BUO -- www.nac.net net @ccess corporation, 201-983-0725 -- 201-983-0725

1 0

Re: Internet Backbone Index
by Robert Laughlin 09 Jul '97

09 Jul '97

In response to the *smart* comments (not repeated here), the percentages given are of external traffic only. The 2% not accounted for is due to rounding down the results. I suppose I can do it to more decimal places if you really want it. Also we do retain lots of our own traffic, but I felt it was irrevalent for the purposes of this discussion, so retained traffic was not included in the calculations. Best Regards, Robert Laughlin ---------------------------------------------------------------------------- DataXchange sales: 800-863-1550 http://www.dx.net Network Operations Center: 703-903-7412 -or- 888-903-7412 ---------------------------------------------------------------------------- > > DataXchange purchases transit from uunet, mci & sprint. But we also > > aggressively try to peer with those who wish to peer. Gary, you are wrong > > to say that 90% of traffic is for uunet, mci or sprint. I have the > > statistics on my own network to prove this. Typically we see a breakdown > > like this: > > > > MCI 22% > > Sprint 7% > > UUNET 21% > > all other (via peering) 48% > > > > In our opinion since only 52% of the traffic has its best routing thru > > uunet, sprint & mci, bi-lateral peering is definately worthwhile. > >

1 0

Re: Keynote/Boardwatch Results
by Eric Germann 09 Jul '97

09 Jul '97

If you really want to test raw performance, why not just enable the chargen service on a router in the core and let him suck that down and measure throughput. No more effective DOS attach than 4 simultaneous chargen sessions on a T1. I suspect a 7513 should be able to spit a few characters per second :) Eric At 11:24 AM 7/9/97 -0700, John Leong wrote: >n Mon, 7 Jul 1997, Jack Rickard wrote: >> I discussed the universal test page >> > suggestion with Gene Shklar this afternoon and we will make it so. > >Universal test page is certainly a [good] step - but it is a step (one >among many) towards [comparative] testing the performance between the >test machine and a *SPECIFIC* web server .... but not the peformance of >the backbone (even though it clearly is a major component of the >result). There are simply too many uncontrolled variables still >outstanding. > >By the way, we have given considerable thought on the attributes of such >standardised test page. ... starting with what is the test about. It >would be a good idea if a sub group of interested parties can get off >line and work on this under some frame work (IETF, NANOG, CAIDA, ... or >just an ad hoc BOF group) and then report the finding back to the >mailing list. > >So, if anyone is interested to work on a standardized web page, send me >mail and I will try to get a seperate discussion list going, do some >serious work ... and reduces the S/N ratio of this list. > >Regards, >John Leong > >

2 1