A company runs an application on EC2 On-Demand Instances in an Auto Scaling group. Peak load occurs at the same time every day. Users report slow performance at the start of peak hours, and performance recovers after 2–3 hours. The company wants the application to perform properly at the start of peak hours. Which solution will meet this requirement?

The correct solution is to configure a scheduled scaling policy. Since peak load occurs at the same time every day, a scheduled scaling policy can proactively launch new instances before the peak traffic begins, ensuring the application has sufficient capacity from the start. This prevents the initial slow performance and the 2-3 hour recovery period. An Application Load Balancer distributes traffic but doesn't solve the underlying capacity shortage. Dynamic scaling policies (based on memory or CPU utilization) are reactive; they launch new instances after the performance degradation has already started, which is precisely what the company wants to avoid.