I think the one big drawback for me is the loss visibility and control for the root operators. As an example, DITL, what value will that have if only subset of queries make it to root servers? Will DNS-OARC have to collect logs from all these loopback authoritative slave recursive? -1 for adoption. > -----Original Message----- > From: DNSOP [mailto:dnsop-boun...@ietf.org] On Behalf Of Paul Hoffman > Sent: November-17-14 11:05 PM > To: Nicholas Weaver > Cc: dnsop > Subject: Re: [DNSOP] Call for Adoption draft-wkumari-dnsop-root-loopback > > On Nov 17, 2014, at 5:50 PM, Nicholas Weaver <nwea...@icsi.berkeley.edu> > wrote: > > Trying to be polite here, but this seems just silly, and the only thing > > really > should be "Don't Bother". > > > > > > Root latency frankly speaking does not matter. Lookups to the root > themselves should be rare, and the responses have very long TTLs (48 hours!). > So this is clearly optimizing something that needs no optimization. > > It's fine if you don't want the WG to adopt this draft, but that second > sentence is > clearly wrong. The third paragraph of the introduction says: > > <t>The primary goal of this design is to provide faster negative > responses to stub resolver queries that contain junk queries. This > design will probably have little effect on getting faster positive > responses to stub resolver for good queries on TLDs, because the data > for those zones is usually long-lived and already in the cache of the > recursive resolver; thus, getting faster positive responses is a > non-goal of this design.</t> > > Lookups to the root for things that don't actually exist in the root happen > all the > time, yes? > > --Paul Hoffman > _______________________________________________ > DNSOP mailing list > DNSOP@ietf.org > https://www.ietf.org/mailman/listinfo/dnsop
_______________________________________________ DNSOP mailing list DNSOP@ietf.org https://www.ietf.org/mailman/listinfo/dnsop
