I've got a situation that I'm not sure if I need to change.
I've got a 3750 stack that's using L3 interfaces. It's configured in Dense mode at the moment. There's an uplink that connects to a Dell stack for our edge switches. The Dell stack has igmp snooping configured, and I'm seeing my groups on the switch and what ports are requesting to join that multicast group. Here's the question:
I also see the group on the 3750 for my source. We had an issue where Windows 7 has a migration utility and it brought the network to its knees. According to MS, Windows 7 uses a multicasting type method to find the requesters to be updated. Anyway, my source that's seen on the core is NOT connected to the core....it's connected to the Dell stack. My concern is that if the core is routing for it, that means that traffic is going through the uplink and back to get to the requesters who are also in the edge switch. In this design, my routers don't know about multicasting because we don't truly have a need for it; I'm just trying to control the multicasts that are coming in from random applications. The core (3750) is configured in Dense, but I'm wondering if I should configure itself as an RP and reconfigure in Sparse mode or if this would even matter?
Awesome! I'm going to set that up and let you know
Now, what happens if I need to route between vlans on the same switch....would I need pim enabled on all SVIs?
If you want to route between vlans on the switch then you would need to enable PIM (sparse or dense mode) on all the L3 SVIs for the vlans that you wanted to route between. And obviously you need to enable "ip multicast-routing" in global config mode.