You have been asked to design a fault-tolerant and scalable web application across three Availability zones. The presentation logic will reside on web servers behind an ELB classic Load Balancer, and the application logic will reside on a set of app servers behind a second load balancer. How should you use Auto Scaling groups?