msdp mesh group

oxzgan

i read
"An MSDP mesh-group is a mechanism for reducing SA (Source-Active) flooding, especially in intra-domain environment. An MSDP mesh-group is a group of MSDP speakers with fully-meshed MSDP connectivity between each other. Any SA messages received from a peer in a mesh-group in NOT forwarded to other peers in the same mesh-group. This reduces SA message flooding and simple peer-RPF flooding."

my question is with out flooding how each peer will know about information such as RP ?

mikej412

oxzgan wrote: »

my question is with out flooding how each peer will know about information such as RP ?

Multicast Source Discovery Protocol SA Filter Recommendations - Cisco Systems

MSDP-SA messages contain (source, group (S,G)) information for rendezvous points (RPs) (called MSDP peers) in Protocol Independent Multicast sparse-mode (PIM-SM) domains. This mechanism allows RPs to learn about multicast sources in remote PIM-SM domains so that they can join those sources if there are local receivers in their own domain.

The MSDP-SA message aren't flooded within the mesh group since the other members should have received the same message, but they will be sent to downstream non-mesh group MSDP peers.

Juniper has nice example (and explanation) here: Configure MSDP Mesh Groups (but if anyone asks you didn't get this link from me ).

When a source-active message is received from a mesh group member, the source-active message is always accepted but is not flooded to other members of the same mesh group. However, the source-active message is flooded to non-mesh group peers or members of other mesh groups.

And then once you're within a specific PIM Sparse Mode domain, Amit's Cisco Zone Blog probably has the clearest explanation I've ever seen that puts it all together: MSDP & its operation - Amit's Cisco Zone

When the MSDP peer/RP receives the new SA message, it determines if there are any other group members within the domain interested in any group described in the (S,G) entry within the SA message i.e. the RP checks for a (*,G) entry with a non-empty outgoing interface list (indicating a group member for that particular group). In this case, the RP creates a (S,G) entry rooted to the data-source present in other PIM domain. This sets up a branch to the source-tree to this domain. As multicast packets arrive at the RP, they are forwarded down its own shared-tree inside the domain. If the members' DR wish to join the source-tree, they have the option to do so according to standard PIM procedures.

The Cisco docs do mention the first part (and I like Figure 1.) -- but then it leaves it to you to you to remember your individual PIM Domain processing.

When an RP for a domain receives an SA message from an MSDP peer, it determines if it has any group members interested in the group the SA message describes. If the (*,G) entry exists with a nonempty outgoing interface list, the domain is interested in the group, and the RP triggers an (S,G) join toward the source.