cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
449
Views
0
Helpful
1
Replies

Multicast problem

dhickey
Level 1
Level 1

Hi all,

We are running into a problem with channel changes on and IPTV network. Each channel is a mulicast address.

I have two 6509 switches, each running 12.2(18)SXF3

Switch 1 (S1) has the multicast source located on it.

Switch 2 (S2) has our customers. For my testing of the issue, my Set top boxes are directly connected to Switch 2. This takes everything else out of the picture as far as a customer connects.

Source

|

|

S1

|

|

S2

|

|

Receiver (STB)

We are using IP Pim dense mode.

There is a routing between S1 and S2...

The problem we can duplicate is when the prune timer (00:02:55 in the below example) hits 00:00:00 and a join is requested at the same time, the group is not joined and the STB goes channel unavailable. We lose the (172.17.15.101, 224.1.3.121), route for a while. It takes a while for this to come back.

(172.17.15.101, 224.1.3.121), 00:52:10/00:02:55, flags: T

Incoming interface: GigabitEthernet2/8, RPF nbr 10.157.1.17, RPF-MFD

Outgoing interface list:

Vlan5, Forward/Dense, 00:44:57/00:00:00, H

This only occurs if no one else has joined that group (ie. no one is watch the channel). Is anyone else is on the channel, the timer never makes it down to 00:00:00.

So on the customer end, they are making channel changes and happen to hit the channel at the timer has hit 00:00:00 and the channel is unavailable..

Any ideas? I hope I explained it good enough.

Thanks

Don Hickey

1 Reply 1

dhickey
Level 1
Level 1

For anyone intersted,

I fixed this problem by changing from Dense mode to sparse mode..

Don

Review Cisco Networking for a $25 gift card