[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
DNSSEC Opt In
- To: namedroppers@ops.ietf.org
- Subject: DNSSEC Opt In
- From: Jerry Scharf <scharf@vix.com>
- Date: Tue, 21 Nov 2000 16:22:53 -0800
- Delivery-date: Tue, 21 Nov 2000 16:33:03 -0800
- Envelope-to: namedroppers-data@psg.com
I had a talk with Mark about the draft, and came to the following point that I
wanted to bring to the group for discussion.
Mark's premise is that the NS list for the secure and non-secure zones is the
same and that "magic happens" in some way to separate the secure and
non-secure queries (we discussed several implementation options for this.)
I think that we could look at changing the NS list sets to be explicit. If
there were separate NS lists for with opt DO and without, then there ther is
no need for the "magic happens" part of the process. It would mean having some
way of telling a resolver which opts are covered by this NS set when passing
the glue back. There is nothing that requires running a zone as opt-in, so we
can punt doing this for the root.
It does bring up a problem. If this is a precedent for deploying other complex
options to DNS, they you get into the combinatorial problem of NS sets. I
believe this is a problem of either the implicit or explicit flavors of this
solution, but it is more clear in the explicit case. Without having some
answer for the combinatorial options problem, this becomes a one time bullet.
jerry
to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.