On 06/ 3/10 08:46 AM, [email protected] wrote:
We can imagine a reserved range of addresses in the DHCP server for
random and static addresses (a well known sink or very simple hosts
without DHCPv6 support). But even in that case a safe DAD would require
a multihop DAD. The new bit would just be an optimisation for each
dedicated range of addresses.
That only help if you also reserve the complement of that set from being
used by claim-and-defend.
Hence you would in essence require a cast-in-stone split of the short
address range into
- assignable by dhcpv6
- claim-and-defend
in a standards track document and everybody better adhere to those rules
or you would have duplicates.
That doesn't sound very flexible, since some environments might want
2^^16 addresses for use by DHCPv6, and it doesn't sound very robust
since some admins might want to configure different ranges of short
addresses for their dhcpv6 servers hence might accidentally configure it
to use the claim-and-defend part of the space.
Erik
_______________________________________________
6lowpan mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/6lowpan
