Le 19/10/2012 17:50, Behcet Sarikaya a écrit :
Hi Mikael,
On Fri, Oct 19, 2012 at 3:08 AM, Mikael Abrahamsson
<swm...@swm.pp.se> wrote:
On Fri, 19 Oct 2012, Alexandru Petrescu wrote:
Comments about the idea in this draft? About the problem?
What is the rationale for duplicating the functionality in
DHCPv6-PD into ND? If code needs to be changed, why can't that code
change be to implement existing standard instead of implementing a
new standard?
It is not just implementing that code in one node. DHCPv6-PD
requires Delegating Router on a DHCP server somewhere and then
Requesting Router on the edge router (there was a proposal to
implement it on a UE).
It is a system support issue, I think.
I tend to agree.
I also work in a perspective where it is hard to say today which
particular feature will future ND and future DHCP support, especially in
vehicular environments. In this sense, intelligent mechanisms may be
needed for an arbitrary vehicle to decide what kind of protocol to use
when facing another vehicle ready to provide it a default route and
maybe a prefix: try first ND, then DHCP... or similar logic. However,
as much as I consider a system issue and appreciate the value of an
algorithm implementing it, it may be the case that no real protocol work
could be documented in an Internet Draft about such system issue.
Just some thoughts.
Alex
Regards,
Behcet
Isn't ND handled by the kernel in a lot of OSes? Does prefix
delegation really belong there?
-- Mikael Abrahamsson email: swm...@swm.pp.se
--------------------------------------------------------------------
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
--------------------------------------------------------------------
