An application runs on EC2 instances behind an Application Load Balancer (ALB) in an Auto Scaling group. The application starts producing errors whenever total requests exceed 100 requests per second. The administrator needs to collect total request counts over a two-week period to find when the threshold was exceeded. What is the appropriate way to gather this data?

The ALB RequestCount metric in CloudWatch directly tracks the total number of requests processed by the Application Load Balancer, making it the most straightforward and accurate way to monitor the application's request load. Setting the time range to two weeks and the period to 1 minute provides the necessary granularity to identify when the 100 requests per second threshold was exceeded over the specified duration. Using CloudWatch metric math to sum per-instance request metrics is less efficient and potentially inaccurate, as the ALB already aggregates this data. Adding custom CloudWatch metrics in the EC2 launch configuration would require application-level changes and would be redundant given the existing ALB metrics. Creating an EventBridge rule for EC2 events is not designed for aggregating application request metrics; it's primarily for reacting to state changes or scheduled events.