Multicast question

unclerico

Hey all, I've got a question in regards to Multicast operation and the operation of RP's in particular.

I've got an extremely simple topology consisting of 2 switches and 3 routers (as seen below):
http://www.box.net/shared/ex0qg9lnbj#1:21145254:222375280

R2610 is configured as the RP with the following command:

R2610(config)# ip pim send-rp-announce loop0 scope 15

R3640 is configured as the Mapping Agent with the following command:

R3640(config)# ip pim send-rp-discover scope 15

and each interface on all routers configed to use sparse-mode.

I am using Tgen as a multicast traffic generator on the laptop hangning off of the 3550 and Mcast as the recipient on two laptops hanging off of the 2950.

In looking at the Multicast routing table on R2620 I notice that the RP for the Multicast Group disappears every now and again for a couple of minutes and then reappears.
Here is the first capture. Notice that the RP address is filled in:

(*, 239.1.1.10), 00:00:39/stopped, RP 192.168.255.254, flags: SJC
Incoming interface: Serial0/1, RPF nbr 192.168.0.5
Outgoing interface list:
FastEthernet0/0, Forward/Sparse, 00:00:39/00:02:20

(192.168.10.5, 239.1.1.10), 00:00:39/00:02:53, flags: JT
Incoming interface: Serial0/1, RPF nbr 192.168.0.5
Outgoing interface list:
FastEthernet0/0, Forward/Sparse, 00:00:39/00:02:20

Here is the second capture. Notice that the RP address is empty:

(*, 239.1.1.10), 00:00:26/stopped, RP 0.0.0.0, flags: DC
Incoming interface: Null, RPF nbr 0.0.0.0
Outgoing interface list:
FastEthernet0/0, Forward/Sparse, 00:00:26/00:02:43

(192.168.10.5, 239.1.1.10), 00:00:26/00:02:52, flags: T
Incoming interface: Serial0/1, RPF nbr 192.168.0.5
Outgoing interface list:
FastEthernet0/0, Forward/Sparse, 00:00:26/00:02:43

Here is the third capture. Notice that the RP address is populated again:

(*, 239.1.1.10), 00:03:25/stopped, RP 192.168.255.254, flags: SJC
Incoming interface: Serial0/1, RPF nbr 192.168.0.5
Outgoing interface list:
FastEthernet0/0, Forward/Sparse, 00:03:25/00:02:44

(192.168.10.5, 239.1.1.10), 00:00:21/00:02:53, flags: JT
Incoming interface: Serial0/1, RPF nbr 192.168.0.5
Outgoing interface list:
FastEthernet0/0, Forward/Sparse, 00:00:21/00:02:44

During the times of these captures no changes to the network have been made and mutlicast traffic continues to flow just fine. I guess I'm just wondering what causes the RP to act like this?

cisco_trooper

My initial thought is that you may be getting an intermittent RPF failure. It might be beneficial to check out your IP routing table at the same time you are checking your Multicast Routing table to see if a change in your IP routing table is the source of the problem. Check that out and let us know what you come up with.

While you're at it you should notice that your multicast route (192.168.10.5, 239.1.1.10) does not appear to be suffering from the RPF failure, which is why your traffic continues to flow uninterrupted.

unclerico

CT, thanks for your response. I hear you on your statement, it may be that is in fact what is happening. It's almost acting like it is configured to use sparse-dense-mode even though it's set to use sparse-mode because the flags go from SJC to DC back to SJC.