> Why have hosts needless generate periodic DHCP traffic when > there is no DHCP server present? True that in DHCPv6 the > impact is much more minor as multicasting is used (in DHCPv4 > all hosts on the network receive the packets because they are > broadcast), but it still seems better to me to avoid > unnecessary traffic.
This is excellent operational reason to keep both the M and O bits. All implementations support the M and O bits being received today it is not a problem and has not caused a problem. DHCPv6 is being deployed as we speak. DHCPv6 or DHCP6 lite can provide prefix delegations. DHCPv6 lite is so we can build lighterweight DHCPv6 not replace DHCPv6. All users I speak with will not turn on stateless in the Enterprise except for wireless and mobility. DHCPv6 permits control by the Enterprise administration and that is the process they know and understand and with all the IPv6 transition items to do they are glad DHCPv6 exists now and that several vendors are shipping products or planning to ship DHCPv6 server products. We should definitely not remove the M and O bits at all in 2461 or 2462, or try to identify use of DHCPv6 or DHCPv6 lite. /jim -------------------------------------------------------------------- IETF IPv6 working group mailing list [EMAIL PROTECTED] Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
