Which solution meets these requirements?

Explanations:

While scheduled scaling can manage traffic during known peak times, it does not dynamically adjust to real-time load changes. This approach may lead to underutilization or overprovisioning of resources outside the scheduled times, compromising high availability.

Simple scaling policies adjust capacity based on defined metrics but lack the responsiveness of more advanced options. This could lead to delays in scaling during sudden traffic spikes, potentially affecting application availability and performance.

A Network Load Balancer (NLB) is typically used for TCP traffic and is less suited for HTTP/HTTPS traffic managed by an Application Load Balancer. Additionally, using a Spot Fleet may introduce instability due to potential instance termination, which is not ideal for high availability requirements.

This option effectively uses an Application Load Balancer (ALB) and Amazon EC2 Auto Scaling with a target tracking scaling policy. Target tracking scales the number of instances automatically based on current load, providing optimal responsiveness to traffic changes, ensuring high availability and performance.