1

Instead of assigning everyone an individual public IP address, could we use a very strong NAT system with thousands of private IPs behind a single public IP? This could reduce public IP demand and still allow internal devices to access the internet. Similar to DNS servers, which handle vast amounts of requests efficiently.

Additionally, internal routers and subnet masks could be utilized to effectively organize and manage the large number of private IPs within the network, enhancing efficiency and security.

1 Answer 1

1

We can and we do.

For lookup efficiency, you'd want to use each public-facing transport-layer port (TCP, UDP) per IP address only a single time. That limits you to roughly 64,000 connections per public IP address per transport protocol. Most large NAT systems use a pool of public IPv4 addresses so they can exceed that.

Of course, there's no terminal reason to not use the destination IP:destination port:source port tuple for lookup. That would limit your NAT translation to 64k connections/session per destination IP:port only.

This could reduce public IP demand and still allow internal devices to access the internet.

That was the whole point of RFCs 1627 (superseded by 1918) and 1631 in 1994. The problem today is that we've run out of IPv4 address for servers as well.

Similar to DNS servers, which handle vast amounts of requests efficiently.

DNS is a stateless protocol. Dynamic NAT isn't - a NAT router needs to keep track of each connection/session which is a whole different ballpark.

4
  • Thank you for highlighting the complexity of managing sessions in large-scale NAT. Could using multiple NAT systems, sharing the load among them, improve efficiency and address scalability concerns? Perhaps utilizing multiple NAT systems in parallel could not only boost efficiency but also enhance reliability by avoiding a single point of failure. Dec 19, 2023 at 18:06
  • 1
    Clustering NAT routers isn't more efficient but can increase scalability and availability.
    – Zac67
    Dec 19, 2023 at 18:28
  • 2
    Just in case Zac67's answer was not specific enough on this point: large scale NAT is in use as described in the original question and has been for many years. In a huge variety of networks from small/home networks to large organizations and even carrier networks, it is used to provide access services for large groups of users who do not need to routinely host accessible services. In large networks it is usually called Carrier Grade NAT. There is little improvement to be had from increasing use of NAT when it comes to IPv4 availability these days, it would not change much. Dec 19, 2023 at 18:32
  • Thank you for your answers. Dec 19, 2023 at 19:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.