A university campus has 80 access switches in one STP domain, experiences slow convergence and high BPDU load, needs deterministic root placement at the distribution layer and wants fewer instances than per-VLAN spanning tree. Which STP scalability and root placement design with RSTP versus MST best fits the requirement?

The correct MST design with two instances and distribution-layer primary and secondary roots fits the 80-switch domain by aggregating many VLANs into two forwarding topologies, sharply lowering BPDU load versus per-VLAN spanning tree while retaining rapid convergence, and deterministic priorities ensure stable diameter, predictable blocking on uplinks, and VLAN-group load sharing across distribution switches. Leaving the same two-instance MST design at default priorities is operationally common but fails determinism because root election then follows MAC address and insertion order, allowing an access switch to win root and stretch paths, increase convergence time, and violate the distribution placement requirement. Using single-topology RSTP with correct distribution roots is widely deployed for fast convergence and deterministic placement but fails the instance requirement because all VLANs share one active topology, forfeiting multi-instance load balancing and leaving a single utilization domain instead of the requested reduced yet split forwarding balance. Creating an individual MST instance for each VLAN with correct distribution roots preserves determinism but fails scalability by recreating per-VLAN control-plane state and configuration burden across 80 switches, directly contradicting the fewer-instances goal. Exam caveat is to separate RSTP rapid convergence from MST VLAN-to-instance scaling, and practical check is to verify identical region name, revision and mapping on all switches plus root placement per instance with show commands.