Which of the following would be the best way to test resiliency in the event of a primary power failure?

Production failover is the best way to test resiliency during a primary power failure because it involves actively switching to backup systems or power sources in a live, operational environment. This directly verifies that redundant systems, such as uninterruptible power supplies (UPS) or generators, function as expected and that critical services remain available without interruption. Parallel processing is a computational technique for speeding up tasks, not a method for testing power failure resiliency. A tabletop exercise is a discussion-based training session that simulates a scenario but does not involve actual system changes or power disruptions. Simulation testing uses models to mimic real-world conditions but doesn't involve the actual production environment or its power infrastructure.