Nat pat and routing?

Question

Network administration

Please explain why nat is needed if there is routing between networks

Answer 1

why do you need nat if there is routing between networks

The initial task that, as far as I know, was called to solve NAT (then still NAT address to address) is the exhaustion of free prefixes ("subnets") in the IPv4 address space. It was decided to allocate several prefixes as "local" and administratively (this is important) prohibit their announcement between autonomous systems. When using NAT, an organization that provides outside access to few or no services may be limited to renting an IPv4 address from its ISP. Without NAT, it would be necessary to allocate a free prefix, given peering limitations (for example, prefixes longer than /24 are often filtered out, i.e. instead of 1 address, a block of 256 must be leased).
PAT is logical to consider as development of NAT. Since most services work using the TCP and UDP protocols, you can further save global addresses by translating not an address to an address, but a tuple (global address, port1) to a tuple (local address, port2), gaining another degree of freedom in terms of multiplexing.
Routing and NAT solve different problems. NAT is used because some networks (local networks, RFC1918) are not allowed to be routed on the global internet. But each of the prefixes involved in NAT must be routed in its own domain, the local prefix - on the local network, the global one - on the Internet.

Answer 2

And how, tell me, will you route the addresses 192.168.0.1 and 8.8.8.8? Without nata.