Just quickly updating you on some more IPv6 stuff I am discovering.
So one of the important things to keep in mind with IPv6 is how they are planning subnet allocation. Its a common question because we have never seen prefixes as big as /64 in IPv4!
So, in IPv6 the vision that Cisco and I believe the IETF has is that your average site subnet willl have a /64 allocation for all the hosts (massive right?) So we have /64 for the "prefix" and /64 for the hosts.
Now, what will tend to happen is an enterprise will be assigned either a /32 or a /48 from there ISP, I am going to go with /32 now as that is the example cisco give.
So you might have for example
This written in non short hand looks like this:
So, the first /32 (2001:CDB1) is your top level prefix, the next
0000 is your "site-level prefix" (i.e. geographical site) (making this a /48) and then finally
the next 0000 is your "subnet-level prefix" (making this a /64)
So thats how they want you to allocate your subnets. Now there is a little bit more to it than that, The IETF also want you to assign your site-level prefixes and number them by altering the LEFT MOST BITS first.
The reason for this is quite simple: over time sites will start to require more and more subnets, if you allocate your site-level prefixes from the right-most bit you will limit the room for growth in this site. (For example, if you needed to increase the size of a site-prefix from /48 to 47 or /46 for example, because you had started with the right-most bits you would need to renumber some sites in order to be able to change the prefix!)
Your actual subnets (your /64's) can be allocated starting from the right most bit as normal.
Confused yet? I am still studying away on the rest of IPv6 and have many unanswered questions such as what subnet to use over a point to point link. As I get more information dear reader: you will too.