Re: [v6ops] draft-ietf-v6ops-slaac-renum and draft-ietf-v6ops-cpe-slaac-renum

[Philip Homburg <[e-mail]>]

>The DHCPv6 PD client doesn't know that the prefix is static, just that it
>has the rights to use that prefix for up to the Valid Lifetime.
>It's presumptuous and foolish of the host to over-promise the use of said
>prefix downstream.

I'm not talking about the DHCPv6 client, but about the RA server component.
The RA server gets as input the prefix with lifetimes from DHCPv6, and
(optionally) a request from the operator of the device to advertise a longer
lifetime. In that case the RA server should honour the user's request.

>> So this draft makes the situation worse for static prefixes to better
>> deal with the unwanted situation of dynamic prefixes. To me that strikes m
>e
>> as completely the wrong way around.
>
>You might have to expand on why you think shorter lifetimes are "worse".

In any more complex setup where some servers are used both internally
and externally, there is a good chance that even internally the server is
referenced using a public DNS name.

So if the internet connection goes down and the static prefix is no longer
available due to short lifetimes, then internal connections suffer.