Re: A common problem with SLAAC in "renumbering" scenarios

[Philip Homburg <[e-mail]>]

> if recommending static IP addressing is an idea that 6man wants to
> push, you'll need to reach out to the security and ops areas to
> get their input on this.  I'm not sure this is an issue that 6man
> can resolve fully.

I like static addresses, but it strikes me as a bad idea to limit ourselves
to static addresses at relatively low protocol levels. 

SLAAC already has plenty of protocol mechanism to deal with changing addresses.
What is left is tying that to DHCP-PD.

I have observed many cases where a host that moves from one network to
another doesn't see a link down event. This is most common with VMs that run on
a laptop. 

This means that hosts need to have a robust way of detecting that the network
environment has changed. There is no reason to assume that routers will have
either unique link local address or unique mac addresses.

At the same time, CPEs that are not cheapest-of-cheapest typically have a 
logfile and record diagnostic information. Adding a bit of flash that lasts
the lifetime of the CPE doesn't cost that much. So there is plenty of room
to specify what a CPE should do.

Whether users get stable prefixes or not should be a matter of policy, not 
something we hardwire in protocols. The latter just leads to situations
where users get told to reboot all of their devices.