Which solution will minimize the failover duration when a network engineer observes a 90-second interruption during a failover event after configuring a second AWS Direct Connect connection and running a test in the AWS Direct Connect Resiliency Toolkit?