North American Network Operators Group Date Prev | Date Next | Date Index | Thread Index | Author Index | Historical Re: Just got on this thing (perhaps very belatedly) - root server trouble?
> At 5:23 PM 2/18/97, Karl Denninger wrote: > [...] > >BTW, churn is the right word. Its taking anywhere from 5-10 *seconds* to > >come back as NXDOMAIN on each request for those that fail to resolve, and > >this is from the IANA roots. > Churn shmurn. Those domains are probably ones that have been paid for (to > the InterNIC), but aren't yet being used. Who's accessing those unused > domains and doing all this needless churning? We should find 'em and > string 'em up. > > Seems like most of the churning would be caused by spammers and testers > like yourself. > > I'd be interested in seeing actual machine statistics on how much > performance degredation can be attributed to lack of responses. Without > those statistics, I can't see how the InterNIC fees aren't covering this > scenario. Well, RFC2010 specifies a latency of 5ms at 1,200 requests/second. I can guarantee you're not meeting that right now on any of the existing COM servers. I'm seeing five *SECOND* response times right now to get back an NXDOMAIN. Lots of people hit non-existant domains. The problem is that this is only a linear degredation problem for a while -- when working sets get into the hundreds of megabytes (as they are for the COM tld servers right now) degredation isn't linear any longer -- its far worse. Pull 60% of the records off those servers and performance would improve by far more than 60% -- it would probably cut average service times by at least 75%, and I wouldn't be surprised to see latencies drop by 90%. > As was mentioned before, you shouldn't have to pay an ISP to have a domain > name reserved. > > Chris Russo Why not? You have to pay NSI! If you're not going to *USE* the domain, why should you be able to register it at all? DNS names aren't things you bandy about - - they exist to perform a translation function. Tell me what the difference is between $50 a year and $100 a year? Not much. You can't get away from the first, and I don't see what the big deal is with the second, given the existance of the first charge. And by the way, from the analysis that I've done, if you think $50 a year is bad from NSI wait until the IAHC's domains come online. With the stats that I have right now on bogus nameservers and domains I'm willing to bet the *break-even* price for those new registrars is going to be closer to $200 a year per domain -- not $50.00. NSI is going to be the *LOW* price supplier under the IAHC proposal. You heard it here first. And the only way to prevent *THAT* is to force free-market competition into the root level of the domain tree. I think we can easily make a profit at half of NSI's fee ($25/year). But there's no way we can do it for $25.00 under the IAHC's plan with the overhead and policy things they're mandating. That's just on the *economic* front. Folks, we run the network (this *IS* NANOG, right? :-) Let's start actually running it for a change... DNS is one of those things that we ought to be able to do right, and do in an open and competitive format. -- -- Karl Denninger ([email protected])| MCSNet - The Finest Internet Connectivity http://www.mcs.net/~karl | T1's from $600 monthly to FULL DS-3 Service | 99 Analog numbers, 77 ISDN, Web servers $75/mo Voice: [+1 312 803-MCS1 x219]| Email to "[email protected]" WWW: http://www.mcs.net/ Fax: [+1 312 803-4929] | 2 FULL DS-3 Internet links; 400Mbps B/W Internal - - - - - - - - - - - - - - - - -
|