Delegating Part Of Class C?

Joseph S D Yao jsdy at cospo.osis.gov
Thu Jul 22 20:31:08 UTC 1999


> Name:    host1.mydomain.com
> Address:  xxx.xxx.xxx.82
> Aliases:  82.xxx.xxx.xxx.in-addr.arpa
> 
> While I'm quite pleased to see this resolved, though the solution seems
> very strange to me. I still can't make sense out of how the 5 octet scheme
> makes this work, or why the lookups come out as they do.

Forget any magic about IP addresses.  Just think of this as a funky
domain.  The only magic is that 'nslookup' "corrects" the lookup of an
IP address (xxx.yyy.zzz.82) into the corresponding REVERSE DNS lookup
(82.zzz.yyy.xxx.in-addr.arpa).

If I looked up skree.klingon.emp, and the klingon.emp name server said
that "skree" had as its CNAME "skree.warbird.klingon.emp", then you
would go and look up the latter name.  But you also have to ask the
same name server about the domain "warbird.klingon.emp", and it might
re-direct you to another server for that domain, which would have the
information you needed.

In this particular case, you looked up 82.zzz.yyy.xxx.in-addr.arpa.
The name server for domain zzz.yyy.xxx.in-addr.arpa [your ISP] told you
that 82.[...]'s canonical name [CNAME] was really in fact
82.80.zzz.yyy.xxx.in-addr.arpa.  It also told you that the name server
for the domain 80.zzz.yyy.xxx.in-addr.arpa was ... your name server.
Zipping over to your name server, you asked in that domain, and you
found the information for 82.80.zzz.yyy.xxx.in-addr.arpa.  Which was
also the information for 82.zzz.yyy.xxx.in-addr.arpa [xxx.yyy.zzz.82],
since they were declared by zzz.yyy.xxx.in-addr.arpa's authoritative
name server to be one and the same thing.

Capish?

ISTM that this is also well described in Albitz & Liu's "DNS and BIND",
3rd Ed., although I don't remember seeing it there myself.

--
Joe Yao				jsdy at cospo.osis.gov - Joseph S. D. Yao
COSPO/OSIS Computer Support					EMT-B
-----------------------------------------------------------------------
This message is not an official statement of COSPO policies.


More information about the bind-users mailing list