Re: wildcard domain serving
Matthew Dempsky <[email protected]> Mon, 12 Apr 2010 23:56:08 -0700
| Newsgroups | gmane.network.djbdns |
|---|---|
| Message-ID | <[email protected]> |
2010/4/12 Colm MacCárthaigh <[email protected]>: > 2010/4/12 Matthew Dempsky <[email protected]>: >> 2010/4/12 Colm MacCárthaigh <[email protected]>: >>> Isn't there a risk that this data could make its way into additional >>> sections you serve? >> >> No, tinydns never includes wildcard records in the additional record section. > > These relevant records provided in the e-mail are not wildcard records; > > Z$tld:a.ns.example.com.:hostmaster.example.com > &$tld::a.ns.example.com:86400 > &$tld::b.ns.example.com:86400 These three lines only define SOA and NS records, which tinydns never includes in the additional records section either. > So if a.ns.example.com and b.ns.example.com have A records defined, it > seems like they would be included. I guess it depends on whether those > records are real, or simply defined by the wildcard. Sorry, I don't understand what the issue you're concerned about is. The records that Jeremy posted will have no effect on how tinydns answers any legitimate queries to his name server. At best we're discussing how tinydns forms responses to illegitimate queries. If you're still concerned, can you try restating the perceived problem more explicitly, perhaps with a concrete example? > They may also discard the entire response as authoritative and cacheable :/ Huh?