You manage a Cloud SQL for PostgreSQL instance located in Region1, which has two read replicas: one in Region2 and another in Region3. An unforeseen event in Region1 necessitates a disaster recovery procedure that involves promoting the read replica in Region2. To guarantee that your application maintains the same database capacity during this transition, what steps should you take to ensure continuity before you reroute the connections?