weird RIP prefix: 5f00::/16

[Harrington, Dan]

Arlie asked:
> I've noticed in a lot of the traffic discussing routes on this list,
> that the length of the prefixes mentioned is quite small.  Is this
> normal?  Is this a good thing?

In general, yes, short prefixes are good, in that they
allow a large number of potential systems to be represented
in a succinct manner.  This is the basis of the CIDR model,
which aims to reduce the number of routing table entries
by aggregating a number of contiguous network addresses behind
a prefix shorter than any of the individual entries would be.

> If people are issued a /24 prefix under
> IPv6, it allocates proportionally as much as allocating /24 in IP4
> address space.

No, not really.  If you've got a 24 bit prefix in IPv4, then
you are 75% (24/32) of the way through the address, and you've
only got 8 bits to number the systems within that /24 network
prefix.  In IPv6, a 24 bit prefix is less than 20% of the way
through the address, and thus can "hide" a much larger number of
networks and systems behind that single entry, as you've still got
104 bits to play with.

> Is there something I don't understand, or are huge chunks of IP6
address
> space being thrown about?

I think you've got it...note that all of the prefixes in use on
the 6bone are as defined in RFC 1897, and represent a fraction of
the total address space defined in RFC 1884.

Dan