On Apr 1, 2013, at 22:56, Toerless Eckert <eck...@cisco.com> wrote: > It is clear that 802.11 is particularily challenged with native L2 > multicast because > they never defined a good resilience scheme as for unicast but so far not > for multicast. > Hopefully this will get fixed sometime.
In a MIMO world, the performance gap between unicast and (indiscriminate) multicast is likely to grow instead of shrink over time. Another little observation: Fixing this at the adaptation layer (i.e., where IP packets are mapped into L2) is suboptimal, because you may not have visibility into the number of member stations per AP. Once you have a few dozen on an AP, it may make sense to switch back to multicast. For that AP! At the adaptation layer, you only see the bridged 802 network, so you don't know how many stations share an AP, and it is difficult to send multicast to one AP and multi-unicast to another. It seems really hard to optimize this. Once you have 802.11 (without any per-AP hacks) in your bridged 802 network, you probably want to do 6085 globally. Pet peeve: We like to make things too "transparent", and then wonder how they break horribly when you no longer can see what you are doing. -> http://en.wikipedia.org/wiki/Leaky_abstraction Grüße, Carsten -------------------------------------------------------------------- IETF IPv6 working group mailing list ipv6@ietf.org Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
