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

[Brian E Carpenter <[e-mail]>]

On 2019-02-07 15:46, Fernando Gont wrote:
> On 6/2/19 16:31, Brian E Carpenter wrote:
> [....]
>>> 3) pushing the smarts down to the host layer and detecting / monitoring 
>>> whether any particular prefix is still valid
>>
>> https://tools.ietf.org/html/rfc6059 ?
> 
> Not really. Actually, the tricky part in this case is that the network
> -- from the host pov -- is the same.

The definition of "the same" is: "The combination of the link-layer address
and the link-local IPv6 address of a router is unique" according to RFC6059.

Maybe a CE router should change its own link-local address when it gets
a new prefix? That would automatically break any attempt to use obsolete
RA information.

    Brian
 
> 
>> In any case, because neither of the preceding options is guaranteed,
>> I think there is no choice: hosts must be able to recover from
>> dead addresses.
> 
> Agreed.
> 
> 
>