A large enterprise customer is migrating thousands of retail offices from legacy TDM circuits to an Ethernet-based service. The network is running OSPF and has been stable for many years. It is now…

The number of offices being unchanged is irrelevant - removing the aggregation layer eliminates the summarization boundary so the same offices now cause far more LSA flooding.

Physical interface count does not determine OSPF LSA propagation scope; the topology change removes the summarization point, which is independent of interface count.

RIB growth is a downstream effect, but the critical OSPF stability concern is LSDB size and flooding behavior, not the RIB directly.

FIB growth is secondary to the explosion in LSDB size and LSA flooding, which are the root causes of OSPF instability in this scenario.

The aggregation routers previously contained and summarized LSA flooding to their local regions; removing them means every link-state update from every retail office now propagates across the entire OSPF domain, vastly increasing flooding traffic and SPF recalculation frequency.

Without summarization at the aggregation layer, each retail office injects its individual LSAs into the backbone; with thousands of offices, the LSDB grows to a size that significantly strains memory and SPF computation on all OSPF routers.