Dynamic NAT vs PAT

zimskiz

What the difference between them ? When should i use Dynamic Nat(w/o overload) or Pat ?


Just an example to see if i get it:

Pat = one private addr mapped to one public addr right?

Dynamic Nat = i have a list of Private Ip's that in mapped to a list of Global IP's
(so if i have in my network 10 PC's and 10 Global IP's everything works fine, but if i add one more pc and all of the 10 PC's are connected to Internet, the last one i ve just added will not be able to reach Internet)

The solution for this problem is Dynamic Nat with overload?

lantech

Do you know what static NAT is?

With Dynamic NAT the translation mappings change so multiple users can use one IP address. But that one IP address can only be used by one user at a time. It's just that once one user is done another user can then use that address.

With PAT multiple users can use one IP address at a time. NAT will translate each IP address to the one address and keep track of each user using available port numbers. So an address might get translated to 193.175.20: 7001 and another would get translated to 193.175.20: 7002.

I'm at work so I don't have a lot of time to explain but I hope that helps at least a little bit.

SecurityThroughObscurity

in other words.
dyn NAT w/ overload = many to many (with port translations)
pat = many to one (with port translations)

lantech

Static Nat - a one to one mapping that doesn't change. It is manually configured.
Dynamic NAT - a one to one mapping that can change. When a device stops using an address it is freed up and returned to the pool of address so another device can come along and use it.
PAT - One address can handle multiple devices. Devices get translated to one address and are assigned a port number that is not in use. The router can then keep track of the device by mapping the port number it gives it to the IP address of the device.

xXErebuS

SecurityThroughObscurity wrote: »

in other words.
dyn NAT w/ overload = many to many (with port translations)
pat = many to one (with port translations)

Dynamic NAT w/ overload is PAT.


lantech I agree with your explanation.

To answer the question you would use PAT when you need more public IP addresses for Private IP addresses than you have available. For example if I have 10 users that constantly go on the "internet" and 8 public addresses, I will use PAT. If only 4 users have access to go on the "internet" I will use dynamic NAT.

Now if those users have something initiating connection to them I will want to use a static NAT since it never changes.

zimskiz

So recap! I know what static NAT means.

PAT = multiple Private addr to one Public addr (with port translations)

Dynamic Nat = multiple Private addr. to multiple Public addr. (but i think it s without port translation 'cause we use only the maximum number of Public addr. that we have)
Dynamic Nat with Overload = how it's done ? we will still use port translations ? if yes, why they make Dynamic Nat with Overload if it's the same thing like PAT?

SecurityThroughObscurity

It is just term.

NetworkVeteran

zimskiz wrote: »

Dynamic Nat with Overload = how it's done ? we will still use port translations ?

Yes, PAT / NAT Overload is associated with port translations.

if yes, why they make Dynamic Nat with Overload if it's the same thing like PAT?

Why is the vehicle in front of my home called both 'car' and 'automobile'? Word origins may be interesting, but both terms refer to the same set of commands on a Cisco device.