Reverse Path Forwarding (RPF) Check is used to prevent multicast routers from sending/receiving duplicate packets. It’s also used to ensure packets are being sent/received on an optimal path. This done by comparing the path of incoming packets with the router’s unicast routing table.

Let’s illustrate with the topology below:

Reverse Path Forwarding (RPF) Check
Reverse Path Forwarding (RPF) Check

There are situations where R1 may send its multicast stream to both R2 and R3. In turn, R2 and R3 would both send the same multicast packets to R4. When R4 realizes that it is receiving duplicates of the multicast, it will perform the Reverse Path Forwarding (RPF) Check. The router will check its unicast routing table to see what path it would choose if it wanted to send traffic to the source. It will accept multicast traffic from this path and drop traffic from the other path. Now the receiver won’t be bothered by the duplicate traffic.

To avoid repeatedly sending unwanted traffic and then needing to drop it, there are mechanisms that can form a distribution tree from the unicast routing table. Distribution Trees are discussed in more detail in the Protocol Independent Multicast lessons.

RFC 5496: The Reverse Path Forwarding (RPF) Vector TLV

To continue reading, please login or become a member for full access...