On Thu, Jun 23, 2005 at 10:03:36AM -0400, Vlad Yasevich wrote: > Tim > > Wouldn't an update to a policy table be enough to resolve the problem?
Yes, but isn't it best that the default policy does the right thing? Else everyone with ULA doing global multicast will need to update their policies. Of course vendors could use their own default policy that is different from 3484's policy... Stig > -vlad > > On Thu, 2005-06-23 at 12:43 +0100, Tim Chown wrote: > > Hi, > > > > As part of our studies in IPv6 renumbering(*) in the 6NET project we've > > been looking at running ULAs alongside globals for internal communication > > stability through a renumbering event. > > > > We seem to have hit an issue with ULAs and multicast, that we'd like to > > get WG comment on. > > > > When a node is multiaddressed with ULA and global prefixes, and wishes > > to send to a globally scoped multicast group, RFC3484 policy will select > > the best match source address of equal scope. The 'snag' is that, as > > per section 3.3 of the ULA draft, ULAs are global scope. As a result, > > multicast traffic to global scope groups will be sent with ULA source > > addresses. > > > > We note that RFC 3879 on site local deprecation doesn't mention multicast, > > we suspect because the deprecation text was written pre-ULAs. > > > > RFC 3484 talks of scope comparisons in section 3.1, but the language of > > 'site locals' is now out of date. > > > > So, if this is an issue, should it be fixed by a recall of ULAs from the > > RFC editor's queue, or an update to RFC 3484? It seems a 3484 update > > is needed at some point due to the 'site local' references in it. > > > > Comments? > > > > > -------------------------------------------------------------------- > IETF IPv6 working group mailing list > ipv6@ietf.org > Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 > -------------------------------------------------------------------- -------------------------------------------------------------------- IETF IPv6 working group mailing list ipv6@ietf.org Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
