> On 30 Jun 2018, at 2:47 am, Paul Vixie <p...@redbarn.org> wrote: > > > > Michael Sheldon wrote: >> I kind of agree it's the long way around. The problem it gets around >> is recursives not returning Additional, which has been noted as being >> very common. > > recursives could (as could authoritatives) do this today if they wanted to. > if you think they will only do it if there's a document telling them to, > then by all means write that document. but just describe what they can > already do, don't define new protocol elements.
It’s not so much that they don’t return the records. It’s that they don’t have the records in the cache to start with so they are not returned. This is especially true when the content is hosted by a different organisation. It is possible to fetch all the records before returning if you want to though some sort of signal from the client that they will use the data would be useful. I’ve just been working on code to do just that for BIND for SRV as a proof of concept. It is also possible to prefetch missing additional records. Both of these things are being done in this MR though I might increase the number of prefetch slots used. https://gitlab.isc.org/isc-projects/bind9/merge_requests/449 > -- > P Vixie > > _______________________________________________ > DNSOP mailing list > DNSOP@ietf.org > https://www.ietf.org/mailman/listinfo/dnsop -- Mark Andrews, ISC 1 Seymour St., Dundas Valley, NSW 2117, Australia PHONE: +61 2 9871 4742 INTERNET: ma...@isc.org _______________________________________________ DNSOP mailing list DNSOP@ietf.org https://www.ietf.org/mailman/listinfo/dnsop
