|
North American Network Operators Group Date Prev | Date Next | Date Index | Thread Index | Author Index | Historical Re: Alternatives (was Re: whois broke again?)
William Allen Simpson wrote: > Trust has to start somewhere. If you, GeekTools, or CenterGate, are > not willing to be trustworthy, then I guess we need to find somewhere > else. Errrr, I didn't mean it to come out that way :-) I think that we're inappropriate in that we are not independent. > > How about Bill Manning at ISI? What Bill Woodcock says :-) > > I use OpenBSD, which has the modifications. I had had high hopes > for whois-servers.net, but it has not solved any problems. Perhaps > I don't understand how it works? whois-servers.net contains pointers to the authoritative whois servers for all the tld's we know. So you would query, for example, using dig, to find out what today's answer is to the question: "What machine is authoritative for a port 43 question inquiring after the whois data for 'domain'" e.g., It does not contain any whois data itself. > What I am suggesting is a set of redundant servers, A.whois-servers.net, > B.whois-servers.net, etc., that mirror each other's data, eliminating > single points of failure. But what data would they contain. whois data for the world? hardly likely to work. Who would run the organization to populate it? > > While I think that Bill Manning's DNS TXT suggestion is clever, and > nicely distributed, it requires a lot of effort. Yes, although someone here condemned it as a hack, I believe that DNS is perfectly positioned to fill the role. Sean believes that we cannot rely on DNS to present the data because it is "in-band". Then let's identify a solution to mirror the data "out-of-band". Nonetheless, I suggest that the correct place to delegate responsibility *is* at the edge. What about SRV? A suggestion has been made that: _whois._tcp.centergate.com IN SRV ... _whois._tcp.5.5.192.in-addr.arpa IN SRV ... is the right way to go. Thoughts? > > I'm suggesting a low effort technique to collect the information that > exists. That is, to use the actual whois searches that are done, > collecting the results in a new database, accessible by existing tools, > or minor modifications of tools. But the data is bad, out of date, useless, etc. > Let's discuss the alternatives, and get busy. Hopefully that's what this is :-) -- Rodney Joffe CenterGate Research Group, LLC. http://www.centergate.com "Technology so advanced, even we don't understand it!"
|