Is IPv6 MLD snooping performed by IOS switches over solicited-node multicast addresses ?

[computerone1]

The NDP protocol makes use of the solicited-node multicast address.

A host joins its own solicited-node multicast address group both localy (netsh int ipv6 show joins) and on the layer2 domain (a wireshark trace shows that a host configured with a new IPv6 address sends a MLD message for its solicited-node multicast address).

Yet, a Cisco layer2 switch (S0) with IPv6 MLD snooping enabled, placed in between three hosts PC1,PC2,Server1 and an IPv6 PIMv2 router shows:

. no display of solicited-node multicast groups

. a recorded link-local multicast group (Multicast DNS, FF02::FB)

Doing some experiments shows that:
. IPv6 MLD snooping is functionning OK with regular IPv6 Multicast traffic, with the traffic correctly switched relevant to the multicast joined clients (ie not-flooded)
. solicited-node multicast is not snooped, but flooded instead

So is IPv6 MLD snooping performed by switches over solicited-node multicast address? Or is it flooded by design ?

Thanks for any input :-)

[Harold Ritter]

According to RFC4451, all link scope (ff02::/16) multicast address should support MLD, except for FF02::1. In principle, MLD should apply to all these multicast groups.

https://www.ietf.org/rfc/rfc4541.txt

RFC4451, section

In IPv6, the data forwarding rules are more straight forward because
   MLD is mandated for addresses with scope 2 (link-scope) or greater.
   The only exception is the address FF02::1 which is the all hosts
   link-scope address for which MLD messages are never sent.  Packets
   with the all hosts link-scope address should be forwarded on all
   ports.

 Regards,

[JCLB]

As each host creates at least one multicast sollicited node address, you would end up with enormous MLD tables.

Read this nanog archive https://archive.nanog.org/sites/default/files/tues.general.wheeler.neighbor.12.pdf