On Mon, 1 Oct 2007, Brian Dickson wrote:
> Ralph Droms wrote:
> > Cisco's CNR and IOS DHCPv6 servers both implement PD. The IOS DHCPv6
> > client implements PD, and can automatically assign /64s from a
> > delegated prefix to downstream interfaces.
> I'm a bit confused by the above. It might be terminology, or your use of
> it, but either way, it would be helpful to give some examples of IOS
> server(s) and IOS client(s), doing DHCPv6 vanilla, and DHCPv6 with PD.
>
> E.g.
> r1 (IOS box, acting as DHCPv6 server)
> r2 (IOS box, acting both as DHCPv6 client to r1, and DHCPv6 server to h1)
I'd assume r2 is dhcpv6 client to r1 and assigns /64s via RA to h1
> h1 (DHCPv6 client)
>
> r2 requests address from r1, gets A1-1 (address 1 from address block A1).
> r2 assigns A1-1 to its interface on-link with r1. r2 sets its default to
> whatever address was given to it by r1.
> r1 delegates prefix A2 to r2.
> r2 assigns A2-1 to itself, and applies A2-1 its downstream interface,
> which is also serving DHCPv6.
> h1 request an address from r2.
> r2 assigns A2-2 to h1. r2 also sets the default gateway to A2-1 so h1
> uses r2 (specifically A2-1) as its default route.
>
> Is this what you're talking about?
>
> If so, what additional assumptions are built in, about the interface
> addresses and block sizes?
> Will assignments only occur if the prefix is /64?
> What happens if the PD is something other than /64? E.g. a /80, or a
> /120, for examples.
>
> Thanks,
>
> Brian
>
> --------------------------------------------------------------------
> IETF IPv6 working group mailing list
> ipv6@ietf.org
> Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
> --------------------------------------------------------------------
>
--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------
