Question 760
An imaginary enterprise networking change reveals: A data center adds server racks but all east-west flows still cross one aggregation switch and congestion grows. What observable result BEST validates the recommended correction?
Explanation
Correct Answer
A data center adds server racks but all east-west flows still cross one aggregation switch and congestion grows. A non-scalable fabric topology concentrates east-west traffic and creates common bottlenecks. Design supported CloudEngine leaf-spine connections with appropriate parallel routed paths and capacity.
Incorrect Answers
A data center adds server racks but all east-west flows still cross one aggregation switch and congestion grows. RDMA-based Ethernet storage may require supported congestion management and low-loss fabric settings to avoid severe performance loss, whereas a non-scalable fabric topology concentrates east-west traffic and creates common bottlenecks.
Controller-driven fabrics can drift if manual device state conflicts with the authoritative design model, while a non-scalable fabric topology concentrates east-west traffic and creates common bottlenecks.
High-availability fabric claims fail when parallel logical paths share an unmodeled physical dependency, but a non-scalable fabric topology concentrates east-west traffic and creates common bottlenecks.