How easy is it for a bogus spider to forge the reverse-resolve lookup results?
I understand that if you have an IP number of A.B.C.D, it gets reverse-resolved by asking some server about D.C.B.A.in-addr.arpa
That's just about the extent of my knowledge. I do know that you can get a forged domain name back (by "forged" I mean that the doman is not registered with the root name servers).
I'm wondering because I'm interested in how bogus spiders are able to fly under the radar. They seem to be able to change their reverse lookup, so that if your httpd logs show reverse-resolved domains, you often don't notice that more than one domain is coming from the same exact IP number. And typically, these bots are polite, and they request robots.txt frequently, even though they may ignore it.
You can get pretty far crawling the web with such a bogus bot, because most webmasters won't notice.
Do you have to involve your upstream provider to feed out bogus reverse-resolves? I'm assuming you have zero need for a real domain name, because you don't have a website, and no one has cause to even use your domain. The whole point is to shift around some bogus domains so that you show up in logs under various disguises, for those who log domains instead of IP numbers.