A three-campus enterprise sees OSPF instability because every access VLAN is advertised across the core, exposing failures and inflating tables. The design must hide access specifics, advertise only aggregates between campuses, and block rogue routes at borders. Which summarization and filtering design best meets this boundary requirement?

The correct design places summarization at the distribution to core boundary so access VLAN flaps are hidden behind stable aggregates, which keeps core link-state databases small, contains failures within the campus, and meets the requirement to advertise only aggregates between campuses, while explicit filtering of specifics at campus borders enforces policy and blocks rogue routes. Permitting all specifics across campus borders while summarizing toward the core fails by concept of border prefix filtering, because without deny rules for more-specifics the aggregates do not prevent leakage and instability propagates. Summarizing only at core routers fails by concept of summarization placement, because access specifics must still traverse distribution to core links before aggregation, so the core remains exposed to churn and table inflation. Summarizing locally but still exchanging full specifics between campuses fails by concept of inter-domain aggregation, because it preserves detailed visibility across the WAN and violates the advertise-only-aggregates requirement despite local summarization. Exam caveat: distinguish ABR summarization, which advertises an aggregate and suppresses the covered specifics at that boundary, from border filtering, which separately blocks rogue or unexpected routes arriving from other campuses. Practical check: inspect the OSPF database and routing table on core and border routers and inject a test access more-specific to confirm only the aggregate is propagated and the specific is denied at the border.