Accton Technology ES4626 - Manual (Page 683)

683
Chapter 17 IPv6 Multicast Protocol
17.1 PIM-DM6
17.1.1 Introduction to PIM-DM6
PIM-DM6（Protocol Independent Multicast, Dense Mode）is the IPv6 version of
Protocol Independent Multicast Dense Mode. It is a Multicast Routing Protocol in dense
mode which adapted to small network. The members of multicast group are relatively
dense under this kind of network environment. There is no difference compared with the
IPv4 version PIM-DM except that the addresses it uses are IPv6 addresses. Thus we
don’t differentiate between PIM-DM and PIM-DM6 in this chapter. All PIM-DM in the text
without specific explanation refers to IPv6 version PIM-DM.
As a result of continuous development of IPv6 network, it has the network
environment of nonsupport IPv6 multicast sometimes, so it needs to do the IPv6
multicast operation by tunnel. Therefore, our PIM-DM6 supports configuration on
configure tunnel, and passes through nonsupport IPv6 multicast network by single cast
packet of IPv4 encapsulation.
The working process of PIM-DM can be summarized as: Neighbor Discovery,
Flooding-Prune, and Graft.
1. Neigh hour Discovery
When PIM-DM router is started at beginning, Hello message is required to discover
neighbors. The network nodes running PIM-DM use Hello message to contact each other.
PIM-DM Hello message is sent periodically.
2. Flooding-Prune
PIM-DM assumes that all hosts on the network are ready to receive multicast data.
When certain multicast source S begins to send data to a multicast group G, after
receiving the multicast packet, the router will make RPF examination first according to
the unicast table. If the check passes, the router will create a (S, G) table item and
forward the multicast packet to all downstream PIM-DM nodes (Flooding). If the RPF
examination fails, i.e. the multicast packet is inputted from the incorrect interface, and
then the message is discarded. After this procedure, every node will create an (S, G)
item in the PIM-DM multicast domain. If there is no multicast group member in the
downstream nodes, then a Prune message is sent to upstream nodes notifying not to
forward data to this multicast group any more. After receiving Prune message, the
corresponding interfaces will be deleted from the output interface list corresponding with