There seems to be some confusion regarding the default behavior of NAT and its configuration.
By default, the purpose of Network Address Translations is to translate the SOURCE address of a packet that may be received by either the "inside" interface or "outside" interface. Either way, it's the SOURCE address that's translated.
The difference between the 2 is reflected in the configuration.
For example, the command
ip nat inside source static 1.1.1.1 2.2.2.2
will inspect the packet received on the "inside" NAT interface and translate the source IP address from 1.1.1.1 to 2.2.2.2.
On the other hand, the command
ip nat outside source static 2.2.2.2 1.1.1.1
will inspect the packet received on the "outside" NAT interface and translate the source address from 2.2.2.2 to 1.1.1.1
Notice that in either case, it is the SOURCE address that is NAT'ed. The difference is in the direction of traffic flow -- either coming from the outside and heading toward the inside (outside-to-inside) or coming from the inside and heading toward the outside (inside-to-outside). That difference in direction is also reflected in the order of operations.
Now, if you would like to perform a special case of NATing the destination address of packets received on the "inside" NAT interface, you would use the following command:
ip nat inside destination { list pool | static }
This command is similar to the source translation command. For dynamic destination translation to make any sense, the pool should be a rotary-type pool.
HTH
Victor
