Configuring NAT Using Overloading

Configuring NAT Using Overloading
Once all IP addresses in a pool have been allocated, any new connection attempts will fail.
So if your ISP allocated you only 13 IP addresses, then only the first 13 users will be able
to access the Internet. Once a NAT entry has expired, the IP address is released back to the
pool and the next user will be able to access the Internet. This doesn’t sound like a very efficient
use of the IP addresses.
Configuring NAT 93
Configuring overloading allows the router to reuse each IP address in the pool. It can do this
because it changes not only the IP address but also the port number. This is what is called Port
Address Translation (PAT) or Network Address and Port Translation (NAPT). The router will
add the layer 4 protocol and port information for each translation entry, which allows more
inside IP addresses to access the outside network than there are IP addresses in the pool.
When tying the access list to the NAT pool with the ip nat inside source list command,
the overload keyword is added to configure the overloading feature. The pool of
addresses can even be just one IP address in size, but it can support approximately 64,000 inside
users using a single layer 4 protocol by varying the outbound port numbers. The following
example shows the commands used for the overloading feature:
Border(config)#ip nat inside source list 12 pool OUTBOUND overload
Border(config)#
What happens if you’re using DHCP on the outbound interface and you don’t know what
the IP address is going to be? You can’t configure a pool of IP addresses if you don’t know the
outside IP address. You can configure the router to use the IP address of the outside interface
as the outgoing NAT address by using the same ip nat inside source command but with
slightly different parameters. You still need to specify the traffic to NAT with the list keyword,
followed by the access list number. But instead of specifying a pool name, you use the
interface keyword followed by the interface name and the overload parameter. The router
will then use the IP address of the interface specified. This can be a static IP address or one that
is acquired from DHCP. The following example uses the IP address of the Ethernet1 interface
instead of a pool of IP addresses:
Border(config)#ip nat inside source list 12 interface ethernet1 overload
Border(config)#
Now let’s talk about a pretty cool feature of NAT called TCP load distribution.