A solutions architect is assessing an application's resilience after a database administrator performed a failover on an Amazon Aurora PostgreSQL writer instance, causing a 3-minute downtime. What is the solution with the least operational overhead to minimize downtime during scaling?