Could someone please clarify some questions I have.
1) In a sparse-mode environment does a multicast source need to send out mutlicasts before a client joins via a igmp requests? Assume that the client is in a remote office and the source in another office connected via layer 3 WAN link.
2) If the source isn't multicasting the data and a client requests it, what is going to happen to the request in a sparse-mode environment? How will the switches in know where to get the data from?
3) I have two switches in a site bound together via a trunk connection, all vlan's allowed, how will multicasts traverse the trunk connection as I can't place the command "ip pim sparse-mode" on a trunk link?
1- no not necessarilly. Source may start before or after the client . Both will work.
2- Client will join its connected routers. Routers will then initiate pim sparse-mode join towards the rendez-vous points.
when the source will stream to stream (later eventually) , the source will also inform the Rendez-vous point wiht PIM (register message)
and from there source will be able to send its traffic to the receivers.
3- you can configure a vlan interface on those switches and run ip pim sparse-mode on each vlan. but you will also need to setup a rendez-vous point.