With some subjects, apparently you need to go over them a few times before you get the Aha! moment, maybe even reading stuff from different authors. IPv6 is definitely one of those subjects for me.
I’ve read about IPv6 having a built-in DHCP mechanism, but no book I’ve read so far had a good explanation. However, Wendell Odom, in one concise table, explained it clearly, succinctly. In Chapter 16 of CCNP ROUTE 642-902 Official Certification Guide Book, table 16-5 says it all. This mechanism is called Stateless Autoconfig, and in sum, here is how it works.
The host uses Neighbor Discovery Protocol and finds the local router, from which they learn the network portion of the subnet they are on, in addition to the default router they should use. In IPv4, if you’ve got this much, you’re practically home free, as you’ve got the network portion of the IP address, the subnet mask, as well as the default gateway. All you need is the host portion of the IP address, and you’re done.
In IPv6, you’re almost done too. Using Stateless Autoconfig, the host portion of the address is generated from the MAC address of the NIC, thus insuring all host addresses are unique. Now all the host has to do is combine the host portion of the IP address to the network portion, and voila! One IPv6 address, complete with prefix (or subnet value) and a default gateway.
Thank you Wendell!!