[Ietf-not43] DNS servers in WHOIS output

[Eric A. Hall]

Andrew Newton wrote:

> But since this is for troubleshooting and nobody is saying to do
> backup resolution out of WHOIS or LDAP (atleast I hope not!), it
> isn't necessary to reproduce the DNS infomation verbatim.  So breaking
> this out into separate NS and A records seems to be overdoing it
> a bit.

So read-only replication is okay. Alright, replication from what then?
Does it replicate from DNS where it is the most valuable for diagnostic
purposes, or does it replicate from the registration database? Although
these should be the same thing they won't always be. Updates which go into
DNS the next day, for example, will have different values in the two
systems. Which one to use?

If we do this, how does the data need to be defined? Clearly we want
information to be programmatically parsable for maximum leverage (not as
an out-of-band way to do DNS, but to allow clients to perform routine
tasks on the data, such as querying the servers for the SOA of the zone,
to see if it actually exists). This requires a schema for DNS RRs, if we
want to do it right.

Also, if we're going to do that, shouldn't we also look at providing
routing information with the IP networks?

-- 
Eric A. Hall                                        http://www.ehsco.com/
Internet Core Protocols          http://www.oreilly.com/catalog/coreprot/