Which two steps must a service provider consider when adding multicast support to the existing MPLS L3VPN? (Choose two.)

While SSM (Source-Specific Multicast) can improve efficiency and reduce PIM state complexity by removing RP dependency, its primary benefit is not 'minimizing the number of group addresses needed in the core,' as customer group addresses are determined by the customer, and SSM merely changes how they are handled.

The default MDT (Multicast Distribution Tree) acts as a single, always-on tunnel carrying all multicast traffic for a particular customer VRF across the MPLS core. By encapsulating and aggregating all customer multicast streams into one core tunnel, it significantly reduces the number of individual multicast routing states (PIM joins, (S,G) or (*,G) entries) that the provider core routers need to maintain for each customer group.

It is not a requirement for all PEs to be configured as RPs (Rendezvous Points) for every customer VRF; typically, a few RPs are strategically placed for scalability and redundancy, and having all PEs as RPs for all VRFs would increase complexity.

If a service provider exclusively relies on the default MDT for all customer multicast traffic, every PE router participating in a given VRF's default MDT will receive all multicast streams for that VRF. This can lead to bandwidth waste as PEs receive traffic for multicast groups for which they have no local receivers, consuming unnecessary resources on the PE and core network segments.

Forwarding all customer multicast traffic exclusively over the default MDT does not optimize PE utilization; instead, it can lead to inefficient utilization and wasted bandwidth because PEs will receive traffic for groups without local receivers, increasing processing load and unnecessary traffic.