Hello Kirshan
Let’s take a look at the diagram from the lesson:
Now imagine that the HQ and Branch devices are switches. They will, both together, create a single broadcast domain/network segment with a subnet of 172.16.1.0/24. As far as those devices are concerned, their Fa0/0 interfaces are directly connected. They know nothing of the intervening network infrastructure, due to the creation of the VPLS. So from the point of view of these switches, they learn the MAC addresses “in the normal way”.
Now how do those MAC addresses get from HQ to Branch and visa versa? Well, the VPLS circuit that is created between the PE routers is negotiated and maintained with the exchange of the MPLS label switching. This mechanism includes TLV encoding within the LDP messages that carry the source and destination MAC addresses of individual hosts on each switch. When the HQ and Branch switches receive those frames, they simply populate their MAC address tables accordingly. MAC information is not maintained by the VPLS circuit, but is simply provided to the switches on each end.
If you want to learn more details about this, you can always do a search for “VPLS MAC address learning” in your favourite search engine and you should find some detailed and informative articles on the subject.
Concerning VPWS, MAC addressing is not exchanged. This is because VPWS, unlike VPLS, is a point to point service which means MAC addressing, or any Layer 2 addressing is not possible and not necessary.
I hope this has been helpful!
Laz