This common
DNS trick is a part of every
DNS geek's knowledge base. It
works by using a
CNAME record on an
in-addr.arpa hostname instead of a
PTR record. The idea being that the owner of the
hostname of the
CNAME record will be able to
manipulate it to his will. This is commonly used for when an
ISP wants to
relinquish only
partial control over an in-addr.arpa
zone (i.e. for
colocate purposes).
Confused? Here's an
example:
Normally, an entry in a reverse resolve zone would look like this:
1.20.168.192.in-addr.arpa. IN PTR targethost.foo.com.
However, an entry in the reverse resolve zone using the CNAME trick looks like this:
1.20.168.192.in-addr.arpa. IN CNAME 192.168.20.1.foo.com.
Moving over to foo.com:
192.168.20.1.foo.com. IN PTR targethost.foo.com.
That way, the owner of foo.com has ultimate control of the reverse resolve by using his own zone. This is opposed to having to get ahold of the owner of the in-addr.arpa zone every time the hostname associated with the IP address (targethost.foo.com and 192.168.20.1, respectively) changes.
Cool, huh?