Cooling redundancy means that loss of an individual component or path does not eliminate the facility's ability to remove the required IT heat load . The configurations that provide that capability are C, E, and F .

A standby pump that assumes water-flow duty when the primary pump becomes unavailable provides component-level redundancy. Independent cooling paths capable of supporting the load after one path fails provide distribution-path redundancy. Likewise, multiple cooling units sized so that the remaining units can serve the space following a unit failure provide equipment-capacity redundancy.

Oracle's current design-readiness guidance requires a mechanical design package containing one-line diagrams, airflow data, CFD analysis, and specifically failure analysis for the cooling infrastructure . The stated purpose is to confirm that the cooling systems can support planned rack loads and failure scenarios. Oracle also requires airflow capacity sufficient to maintain supported IT inlet conditions across the rack row.

A describes ordinary operation rather than redundancy. B retains a single cooling path and shared distribution equipment, creating common dependencies. D describes equipment serving different spaces rather than clearly establishing capacity after a failure. The key test is whether an alternate component, unit, or path can maintain cooling when the primary resource becomes unavailable.

===============