A state university runs PIM sparse-mode for lecture capture across a main campus and two remote sites, needing automatic RP failover without manual config on every router and without depending on a single advertisement source. Which rendezvous point discovery design should the designer choose for resilient operation?

Resilient PIM sparse-mode RP discovery in this multi-campus design is best served by Bootstrap Router with geographically diverse candidate RPs because BSR floods candidate-RP advertisements and bootstrap messages so every router dynamically elects the active RP from the same mapping set and reconverges if an RP or BSR path fails, satisfying both automatic learning and avoidance of a single advertisement source. Anycast-RP with MSDP peering is a legitimate and widely deployed redundancy technique for fast failover and shared state, yet it fails the stated constraint because the common RP address must still be statically defined on every router, retaining manual configuration overhead and operational risk during changes. Centralizing BSR and candidate RP at the main-campus core also provides dynamic mapping distribution, yet it conceptually reintroduces a single-site advertisement dependency so a WAN partition or core failure leaves remote sites without fresh RP information despite BSR mechanics. Source-Specific Multicast with IGMPv3 is operationally sound for efficient one-to-many flows using shortest-path trees without a shared tree, yet it fails the factual premise because it is a different forwarding model that assumes known sources and capable receivers and applications rather than providing RP discovery for Any-Source sparse-mode groups. For exam purposes, distinguish dynamic election from static resilience and verify whether the stem emphasizes zero per-router config versus fastest convergence, and in practice validate with show ip pim rp mapping and show ip pim bsr-router plus a controlled RP shutdown to confirm remote sites retain an active mapping.