Multicast on same VLAN

[Cooky]

I have a need to set up something very similar to what jlazzaro has here

We have a 2821 as active L3 device, w/ a 4006 and a bunch of 2950's as core & access switches.
If my multicast source (an Altiris server) resides in the same VLAN as the receiving hosts, do I still need to set up multicast on the router? or will the switches take care of it themselves?

Sorry; really new in the multicast area...
I've read some Cisco docs, but they all seem to pertain to doing multicast w/ multiple routers, and the source & receivers are in different subnets.

 

[jlazzaro]

phew, dug that one up from the depths didnt you?

if all mcast traffic is limited to a single vlan / broadcast domain, you shouldnt need to touch the router. igmp on the switches will take care of the "flood gates".

think of it (in sorts) like DHCP. if you have a DHCP server with clients on a single broadcast domain, do you need helper-addresses on the router?

 

[Cooky]

thanks for the quick reply.
The ip helper example helps.

One more question though:
Every time our server guy drops Altiris images to PC's through multicast, it saturates our network to a crawl, and we've had to ask him to use unicast instead.
By using unicast however, he claims it takes forever to finish each image drop.

By default our 2950's have IGMP snooping enabled:
Global IGMP Snooping configuration:
-----------------------------------
IGMP snooping : Enabled
IGMPv3 snooping (minimal) : Enabled
Report suppression : Enabled
TCN solicit query : Disabled
TCN flood query count : 2
Last member query interval : 1000

What are we doing wrong (or not doing) to handle imaging through multicast?
The imaging server NIC is in the same VLAN as the clients.

 

[spidey07]

If you have igmp snooping enabled the other ports should not receive the multicast traffic and you shouldn't have any performance problems. A sniffer is what you need to really diagnose what is going on here, have it attached to a port that shouldn't be receiveing mcast.

The other thing you could be running into is the switch is flooding the traffic to all ports. This can happen if you have not setup spanning-tree properly, make sure you're using portfast on any switch port that isn't connected to a switch. Otherwise anytime there is a link change (TCN) spanning-tree will lower the mac aging tables and you can very well have a switch that is flooding traffic. You could also be getting TCNs from other switches/bridges.