Mapping class D to Ethernet physical address

Yes. And the host joining a multicast does the same thing, setting its Ethernet interface to receive frames from the MAC address

Bert

----------------------------------------------------------------------------=

--------------

Thanks Mr Bert

Ok , Is the multicating related enteries which made in layer 2 switch are always be static or it may be aging bit set , or... the behaviour of enteries for multicast table in layer 2 switch is depand on the behaviour or say which type of register support by a switch controller (either it static or dynamic)

please tell me.......

Are you talking about IGMP snooping here? If yes, the forwarding of multicasts to a port times out, as stated in RFC 4541, bottom of page

1. It is also dynamic, though, listening for IGMP leave messages.

Have you read through this yet?

------------------------------------

No i m taking about the enteries made in layer 2 switch in its multicast table is what , either it static or dynamic.

Vikrant

What entries?

Take a normal, non-snooping L2 switch. A MAC frame with multicast MAC DA arrives at the switch. The default behavior is what? It is to flood that multicast frame to all active ports. No tables needed.

What node created that MAC multicast DA to begin with?

It was the host that originally transmitted the frame. If that host was sending out multicast IP packets, *it* would be the box that maps the IP Class D to the appropriate MAC multicast DA.

Does the L2 switch know anything about the IP DA, assuming the MAC frame is carrying IP? No, because I said at the top that this was a normal, non-snooping L2 switch. It knows nothing about Layer 3.

So are there any multicast table entries at all? Only if the switch does either IGMP snooping, or GMRP.

If the L2 switch does IGMP snooping, then read RFC 4541 to see about timeout timers.

Bert