draft-ietf-ipngwg-dns-lookups-04.txt

[Mark.Kasper at icn.siemens.com]

Anyone care to answer this guys question?



"Matt Crawford" <crawdad at fnal.gov> on 10/14/99 09:52:01 AM

To:   ipng at sunroof.eng.sun.com, namedroppers at internic.net
cc:    (bcc: Mark Kasper/Service/ICN)
Subject:  Re: draft-ietf-ipngwg-dns-lookups-04.txt




Francis dit:
> => about ordering, something is not specified: when an address is
> searched for a not fully qualified domain name then there is a search
> for IPv4/A and IPv6/AAAA/A6 and in the search list (the list of domains,
> usually initialised with the local domain name).
> Should we do IP version or domain first?

Let me restate the situation.  The user refers to short-form name H,
and the resolver is configured to search domains D1, ..., Dn.

As I see it, the FQDN the user is implicitly naming is the first one
which *exists* from the sequence H.D1, ..., H.Dn.  Then the resolver
gives back whatever suitable records are listed for that FQDN, or
some sort of error if there are none.

> I believe many implementations do domain first, ie. for instance
> look for an AAAA RR for the name completed with search list
> elements then if nothing is found try the same thing for an A RR.

This isn't the same as what I presumed above.  Which more closely
matches the behavior of existing implementations?  If in a v4-only
world I type "telnet H" and H.D1 exists but has no A record, but H.D2
has an A record, what happens?

                    Matt
--------------------------------------------------------------------
IETF IPng Working Group Mailing List
IPng Home Page:                      http://playground.sun.com/ipng
FTP archive:                      ftp://playground.sun.com/pub/ipng
Direct all administrative requests to majordomo at sunroof.eng.sun.com
--------------------------------------------------------------------