Network Address Translation (NAT)

I thought I understood this already, but some discussion with some colleagues is casting some doubt on the subject.

Here is my question:

For example: If you are given 6 public IP addresses, and 30 private IP addresses, and asked to configure NAT. What would work better; Dynamic NAT with overload or "Dynamic without overload"?

I believe that if you used overload in this case, the NAT router would just use one of the public IP addresses for all the private IPs.

Also, if you use "dynamic", what happens if all the 6 IPs get exhausted - the remaining 24 hosts wouldn't have access to the internet.

What do you do in this situation?

My understanding is that PAT or overloading is defined as many to one. With address overloading many private IP addresses can access the Internet with one global address.

With dynamic we would be talking about a defined range
of addresses on both inside and outside

Hi, Thanks - I was certain that with many-to-one PAT is the solution, but concerned about handful of public IPs and several private IPs. I guess the only viable solution would be to use pooled NAT (even though there are chances of exhausting the public IPs).

Thanks again.