From: Hobbs (deadheadblues@gmail.com)
Date: Wed Jan 14 2009 - 20:50:33 ARST
hello,
I am trying to learn more about bidirectional PIM but no matter what type of
lab topology I come up with, multicast works anyway. Anyone know a good
topology I can configure where multicast would fail unless you enable bidir
pim?
The topology I use right now has this (copied from doccd config guide)
R1 (Receiver) ---- R2 ---- R3 (RP) ----- R4 ----- R5 ----- R6 (Sender)
|
|
R7----- R8 (Receiver)
If you cant see the topology correctly R4 is where the split to R7 happens.
R1 and R8 are receivers
R3 is the RP
R6 is the sender.
All interfaces are PIM-SM, Autorp listener everywhere
Even with this setup, R4 will accept (S,G) incoming from R5, forward it to
R3, then accept (*,G) on that same interface from R3 to forward it out
R7/R8. I always thought this could not happen. I have disabled the
SPT-switchover (ip pim spt-threshold infinity).
R4#sho ip mroute 225.0.0.1 | begin \(\*
(*, 225.0.0.1), 00:12:10/00:03:08, RP 3.3.3.3, flags: S
Incoming interface: Serial1/0, RPF nbr 192.168.34.3
Outgoing interface list:
Serial1/1, Forward/Sparse, 00:12:10/00:03:08
(192.168.56.6, 225.0.0.1), 00:00:22/00:03:21, flags: T
Incoming interface: Serial1/2, RPF nbr 192.168.45.5
Outgoing interface list:
Serial1/0, Forward/Sparse, 00:00:22/00:03:09, A
Serial1/1, Forward/Sparse, 00:00:22/00:03:08
On R4:
s1/0 = R3
s1/1 = R7
s1/2 = R5
thanks,
Blogs and organic groups at http://www.ccie.net
This archive was generated by hypermail 2.1.4 : Sun Mar 01 2009 - 09:43:38 ARST