> > > Updating 3484 might be possible. Something like: pick a random one if > > > some of the addreses turn out to be equivalent? > > > > Yeah, that's the kind of thing that sprang to my mind. > > Or rather: use a sort where the order of equivalent records is > randomized. Dunno if there's an accepted term for that, maybe a > forced unstable sort?
Note that RFC 6724 has this: Obsoletes: 3484 Let's concetrated on updating the right RFC. The problem usually is the following: Rule 9: Use longest matching prefix. When DA and DB belong to the same address family (both are IPv6 or both are IPv4): If CommonPrefixLen(Source(DA), DA) > CommonPrefixLen(Source(DB), DB), then prefer DA. Similarly, if CommonPrefixLen(Source(DA), DA) < CommonPrefixLen(Source(DB), DB), then prefer DB. For IPv6, CommonPrefixLen is limited to 64 bits. So if the AAAA addressses are on the same subnet then Rule 9 doesn't apply. But in all other cases this rule is likely to prefer one specific address. So I guess Rule 9 and 10 need to be augmented with something like: if the source is DNS, then randomize the order of the addresses. _______________________________________________ DNSOP mailing list -- dnsop@ietf.org To unsubscribe send an email to dnsop-le...@ietf.org
