Le 20/12/2012 14:02, Randy Bush a écrit :
"The IID consists of N bits that have no meaning; the only
constraint
Hmm.. how would this work with RFC5453 reserved IID space we
already have for anycast addresses?
is anyone aware of any deployment of the ipv6 invented anycast?
In Mobile IPv6 the HA discovery (DHAAD) uses a dedicated anycast address
valid within the subnet of the HA. This is implemented and demonstrated
although not sure about the scale of deployment.
like most ipv6 magic, i think it is ignored and regular old
ipv4-style anycast is used.
The anycast related to MIP6 works without any IPv4.
Some magic of IPv6 anycast _is_ ignored, in that the IPv6 anycast
address is treated like any other IPv6 unicast address - it is assigned
on an interface, ND is performed on it, etc. It's just that it has a
particular 'reserved' anycast meaning in the head of the implementer,
not that routing code is modified.
AS the MN does not know the address of its HA, it must discover it
within its home network, so it sends a request to a 'reserved' address
which has the prefix of the home network, and a 'reserved' IID - RFC2526
'Mobile IPv6 Home-Agents anycast'.
It could as well have been 'multicast', as multicast is used in DHCP
discovery phase, or other protocols' discovery phases.
(maybe that's what's needed in any new protocol's discovery phases).
Alex
randy
--------------------------------------------------------------------
IETF IPv6 working group mailing list ipv6@ietf.org Administrative
Requests: https://www.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------
--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------
