For the sake of things I said assigned and said that this assignment plays the 
role that was expected from DHCP but not really delivered of getting temporary 
addresses. 

The rest is my French: Using DHCP to assign them is interesting but beyond my 
intent.

Regards,

Pascal

> Le 19 juil. 2016 à 16:05, Michael Richardson <[email protected]> a écrit :
> 
> 
> Pascal spoke today in 6man concerning the process of providing privacy
> enhanced addresses in 6man, and how LLNs typically use their 2-byte short
> addresses for layer-2, and also for forming L3 addresses.
> Pascal mentioned that 2-byte addresses can be dhcpv6 assigned and therefore
> unrelated to the EUI-64.  (Of course, many just use the bottom two bytes of
> the EUI-64, and don't do DAD or ND at all.. Let's agree they are not to spec)
> 
> I was thinking about the dhcpv6 process.  In order to do it in a route-over
> mesh, the route-over mesh needs to be up in order to get traffic to the
> DHCPv6 server.  That means that the DAG has been formed using EUI-64 derived
> link local addresses.
> 
> Afterward, the 6LNs will allocate a 2-byte v6 address from the dhcpv6 server,
> and will then set their L2 address based upon that.    I think that this is
> not specified anywhere...?
> 
> But, the thing that got to write this email (while doing my IETF96 laundry),
> is the parent selection process.  Do we need to have a way for an RPL parent
> to say in it's DIO, "I know you see me as fe80::1234, but you knew me before
> as fe80::1234:56fe:ff78:abcd", such that there could be a seamless and
> efficient transition to the new 2-byte addresses?
> 
> Perhaps it's good enough that the node, having been allocated a 2-byte L2
> address, does a gratitous NA where it announces it's EUI64 LL address with
> it's 2-byte L2 address?  How does that sit security-wise?  Have we just
> encouraged the network to accept gratuitous L2 address spoofing?
> 
> How does this interact with Back-Bone Router EARO processing?
> 
> --
> Michael Richardson <[email protected]>, Sandelman Software Works
> -= IPv6 IoT consulting =-
> 
> 
> 
> _______________________________________________
> 6lo mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/6lo

_______________________________________________
6lo mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/6lo

Reply via email to