I've been meaning to share a summary of December's discussions as they
seemed to me, so here it is:
The problem is extending listing protocol to support listing of a range as
well as a point, hopefully without sacrificing too many data bits.
The current DNSxL protocol SHOULD support ranges of no-data information,
given properly implemented NXDOMAIN handling and presentation of the
appropriate glue.
Alternate listing protocols (such as the ones I presented, and Levine's
b-tree proposal, which could easily extend to other information domains that
can be represented as a b-tree, which is, any mapping with sortable keys,
such as (not to encourage black-hat behavior, but it's a good example of
data as such) a rainbow table)
I pursued a way of packing queries using natural bit flipping instead of
explicit leaf ranges until it became clear that my end result was equivalent
to rfc5782 representation of ipv6 address names, with the addition of
supporting shorter names for ranges.
This resulted in my envisioning what could be a protocol to compete with
draft-levine-iprangepub-01<http://tools.ietf.org/html/draft-levine-iprangepub-01>
in
this space, amounting to a couple of extensions to the 5782 protocol.
In short, when there is a range at a.b.c.rangelister.example, all queries in
subdomains of the range-representing entry, such as
t.m.i.a.b.c.rangelister.example, get responded to with an NXDOMAIN error,
while the range-representing entry holds the data for the range.
This means that a list query library would start with the full query and
work its way up until it gets a result, and the NXDOMAIN at the appropriate
part of the glue determination process will prevent the too-long queries
from reaching the server, as they will be handled by the cache server.
As a policy enforcement measure, a server should identify query sources not
respecting NXDOMAIN responses and block them.
The above is intended as a "representing of ranges" extension to RFC 5782.
On Wed, Dec 29, 2010 at 10:48 AM, John R. Levine <johnl(_at_)iecc(_dot_)com>
wrote:
I've done another version of it, with mostly editorial changes. Comments
as always welcome. I guess now I should implement it.
http://tools.ietf.org/html/draft-levine-iprangepub-01
--
“The aeroplane is fatally defective. It is merely a toy—a sporting
play-thing. It can never become commercially practical." -- Nikola Tesla
_______________________________________________
Asrg mailing list
Asrg(_at_)irtf(_dot_)org
http://www.irtf.org/mailman/listinfo/asrg