I am aware of Levine's draft and it makes more sense than most things I've read on this. But that's putting the cart a bit before the horse in this discussion - which is about trying to ensure that spammers don't get /32 and larger v6 netblocks with the same ease they're acquiring /15s and /16s. As for IPs behind carrier grade NAT - just how many of those do you see operating smtp servers? :) thanks --srs On Wed, Mar 9, 2011 at 3:44 PM, Alessandro Vesely <vesely@tana.it> wrote:
Yes, it is obviously possible. But discovering the right granularity and transmitting it to clients is problematic. For a possible solution see http://tools.ietf.org/html/draft-levine-iprangepub .
OTOH, IPv4 extensively uses NAT, so that blocking the wrong IP may affect an entire network behind it. Traditional DNSBLs would be safer without such widespread use of NAT, which can be taken down using IPv6.
-- Suresh Ramasubramanian (ops.lists@gmail.com)