Hi Bob, inline
On 6 July 2016 at 01:17, Bob Harold <rharo...@umich.edu> wrote: > > On Fri, Jul 1, 2016 at 3:51 AM, IETF Secretariat < > ietf-secretariat-re...@ietf.org> wrote: > >> >> The DNSOP WG has placed draft-song-dns-wireformat-http in state >> Candidate for WG Adoption (entered by Tim Wicinski) >> >> The document is available at >> https://datatracker.ietf.org/doc/draft-song-dns-wireformat-http/ >> >> >> I support adoption and will review. > Thanks > > Is there a need to mention a third scenario? > > Scenario 1: > stub resolver <--- dns or http(s) ---> DNS resolver > Yes. There may be one piece missing is that if http(s) connection is failed (may happen.), the stub resolver should fall back to normal dns without the benefit of this protocol. > > Scenario 2: > stub resolver <--- dns ---> proxy DNS to HTTP(S) <--- http(s) ---> DNS > resolver > Actually the Scenario 2 is : stub resolver <--- dns ---> proxy client <---- dns over http(s) --->proxy server <--- dns --->DNS resolver The proxy client and proxy server in this case are separated severs. There is a note following that proxy client can be implemented in the stub resolver in another process listen to , and the proxy server can be implemented in the DNS resolver in this way. Scenario 3: > stub resolver <--- http(s) ---> proxy HTTP(S) to DNS <--- dns ---> DNS > resolver > > Actually there is a txt saying after the second scenarios which I think includes your proposed case : "It is possible that these scenarios are mixed. The server may speak DNS over HTTP directly and the client use a proxy, or the other way around." > Or is the third a subset of another or implied or not useful? > Some comment on this scenario might be helpful. > Useful but it is treated as a mixed scenarios 1 and 2. One expected audience of this document is the software developer. Scenario 1 requires coder to implement the this protocol in DNS software. The proxy approach is not. > -- > Bob Harold > > > Davey
_______________________________________________ DNSOP mailing list DNSOP@ietf.org https://www.ietf.org/mailman/listinfo/dnsop
