Re: [dnsext] draft-li-dnsext-ipv4-ipv6 -- a more versatile proposal

[Andras Salamon <andras@dns.net>]

On Fri, Nov 06, 2009 at 10:02:05PM +0100, Alfred Hönes wrote:
> -  The RGMASK EDNS option is echoed back (iff the answerer understands
> RGRP and this option).  Any bit still set in the echoed bit mask
> indicates knowledge at the answerer of the corresponding RRset (with no
> such RR present in the answer signalling per-RRtype selective NODATA);
> clearing of a bit would indicate lack of knowledge of the corresponding
> RRset and only be allowed only for cache servers, not for authorities.

The meaning of set/cleared bits is unclear to me.

What is the right thing to do if:
1. a server receives one of these blob queries for type X and Y,
2. the cache contains an authoritative entry for type X,
3. the cache contains an NXDOMAIN entry for type Y?

Due to negative caching there seem to be three possible states:
present, absent, negative-present (this last being a weak proxy
for known-absent-from-zonefile).  Further, does "knowledge" mean
"authoritative" or "present in cache"?

-- Andras Salamon                   andras@dns.net