bidirectional PIM is useful for networks that have many sources and receivers
talking to each other simultaneously and where each participant can become both
the source and receiver of multicast traffic, such as in videoconferencing,
Webex meetings, and group chat. When PIM bidirectional mode is used, the RP
only creates the (*,G) entry for the shared tree. There is no (S,G) entry. This
conserves resources on the RP because state tables for each (S,G) entry are not
In PIM sparse mode,
traffic only flows down the shared tree. In PIM bidirectional mode, traffic
flows up and down the shared tree.
mode also does not use the PIM register/register-stop mechanism to register
sources to the RP. Each source can begin sending to the source at any time.
When the multicast packets arrive at the RP, they are forwarded down the shared
tree (if there are receivers) or dropped (when there are no receivers).
However, there is no way for the RP to tell the source to stop sending
Design-wise you must
think about where to place the RP in your network because it should be
somewhere in the middle between the sources and receivers in the network.
mode has no Reverse Path Forwarding (RPF) check. Instead it uses the concept of
a Designated Forwarder (DF) to prevent loops. This DF is the only router on the
segment that is allowed to send multicast traffic to the RP. If there is only
one router per segment that forwards multicast traffic, there will be no loops.
The DF is chosen using the following mechanism:
The router with
the lowest metric to the RP is the DF.
If the metric is
equal, then the router with the highest IP address becomes the DF.