Simon

Cumbria

IT

A lack of understanding ?

I am absolutely no expert in IP6, but I was under the impression that addresses are going to have a heirarchical structure as a way of minimising the routing table issues that affect the IP4 world.

As it was once explained to me ...
In effect there will be a relatively small number of top level aggregators with 'ownership' of the first few bits of their address space (I assume these will be the large 'clearing houses' like Linx in London). Underneath these will be another level who will have the first few bits from the parent plus some address bits of their own. This will continue down the chain until the end user/organisation gets a block of addresses where the prefix is formed by the chain by which it was reached.

If someone has connectivity from two (or more) different sources, then they will get two (or more) sets of IP6 addresses - and IP6 devices are designed from the ground up on the expectation of having several addresses.

As I interpret what I was told, the routing will be done at a higher level in the chain - with the IP address effectively forming part of the route (up the chain, across to a top level agregator, and down the other side). I imagine that there will be routing between agregators at the lext level, but this will be a LOT simpler than the fairly numerous small blocks currently handled by the worldwide routing tables.

All this also means that whenever someone changes service providers they will also change IP6 addresses - which is more or less the case now for anyone but the very biggest of users, ie all those forced to use addresses from their suppliers allocation. I don't see why this should be a big deal, after all, few people are going to memorise IP6 addresses, so it only requires proper management of DNS.