On Tue, 2005-03-01 at 13:53 +0100, Alexandru Petrescu wrote: >Greg Daley wrote: >> I think that Alex was thinking about end hosts which don't support >> IPv6. > >YEs, hosts that support v4 ok and can't support v6 because of the 33 >requirement of RA (RAs must be sent to 33:33:0:0:0:1 by this RFC). >Receiving an initial RA is essential to configure the stack. Ok, IPv6 >stack configuration could be done manually, but in that case using IPv4 >is more advantageous than using IPv6 because IPv4 uses DHCP >autoconfiguration.
Never heared of IPv6 DHCP ? :) >> It is intentional though that Routers use multicast for Neighbour >> Discovery in order to identify packet destinations, rather than rely >> on flooding. > >Good. It's just that routers should not use multicast for ND when >multicast is not available and simply use broadcast. It's better to use >broadcast and have things working, better than just assuming that >multicast works and have nothing working. If you have such old hardware why even bother upgrading it to IPv6, the OS running on it won't support it either. Ethernet cards go for around $15 US, why not simply swap the card when you are going to do multicast ethernet. As mentioned the driver of the card could solve this itself, if it doesn't upgrade it. >Ok, so that's an option. Another is to just say that when 33 multicast >is not available for broadcast, just use ff broadcast. Which is what various drivers already do. Most of the time when you have a card and RA's don't work, ifconfig eth0 promisc and done. But it is more likely better to upgrade the card as it is prone to have more problems anyway, see above and the other messages. Greets, Jeroen
signature.asc
Description: This is a digitally signed message part
-------------------------------------------------------------------- IETF IPv6 working group mailing list ipv6@ietf.org Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
