Re: Comments on draft-yourtchenko-colitti-nd-reduce-multicast

[Erik Nordmark <[e-mail]>]

On 2/26/14 3:31 AM, Ole Troan wrote:
>
> there is work apparently on failover:
> RFC7031
> draft-ietf-dhc-dhcpv6-failover-design

Looking at those docs, they design seems to focus on address assignment 
using a pair of servers. For instance,

    A subset of available resources (addresses or prefixes) is reserved
    for secondary server use.  This is required for handling a case where
    both servers are able to communicate with clients, but unable to
    communicate with each other.

That is insufficient for address registration, since AR assumes it 
should work unless somebody else has registered the address. Thus this 
would typically require a quorum (2 out of 3, 3 out of 5, etc) so ensure 
that there can be no duplicate registrations when some of the servers 
are down or unreachable.

Thus address registration is subtle different than address assignment, 
and the design center for DHCP is the latter.
(But still investigating what can be done using DHCP.)

    Erik