> On 24.3.2015, at 14.00, Brian E Carpenter <[email protected]> wrote:
> 1. Warn users that renumbering is planned/has started.
> (because long-living sessions will be affected, even in make-before-break)

I am not sure this is really useful - ‘red alert, ISP is about to renumber!’? 
Although, for _general IPv6 case_, perhaps some advanced^2 IPv6 socket API 
might have notification on this as well as e.g. missing lifetime information 
from address information etc.

Speaking of which, cross-platform IPv6 APIs are trainwreck currently if you 
care about e.g. preferred+valid lifetimes of addresses.

> 2. We assume that a prefix delegation or withdrawal from above
> by DHCPv6-PD will trigger the appropriate actions by
> draft-ietf-homenet-prefix-assignment. But I can't tell from the draft
> how that happens. Presumably some process in the relevant CPE does it.

Final withdrawal should just remove the DP (=‘remove ISP’). Before that, new DP 
with better pref value should be available..

> 4. Hosts need new DNS server addresses. I'm not sure who causes
> that to happen.

If they happen to support reconfigure, we could push it over that. It can also 
come via RA, or plain old DHCP will work (because it is NATted address and not 
affected by renumberings). I guess DHCPv6 without reconfigure can be considered 
harmful in this case..

(Or we could just give ULA address of the first-hop router always, and not 
renumber the ULA. *wink*)

Cheers,

-Markus
_______________________________________________
homenet mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/homenet

Reply via email to