You deploy an application across Amazon Linux EC2 instances and need to monitor system metrics fleet-wide. You want to observe the relationship between network traffic and the application's memory usage, with metrics collected at 60-second intervals. Which approach meets these requirements?

The correct option effectively combines AWS services to meet the requirements. CloudWatch detailed monitoring (at 1-minute intervals) is necessary to collect the NetworkIn metric from EC2 instances at the desired 60-second granularity. Basic monitoring only provides 5-minute intervals. The MemoryBytesUsed metric (or memused) is a custom metric that is not natively collected by CloudWatch for EC2 instances; it requires the CloudWatch agent to be installed on the instances to collect and push this data to CloudWatch. Once both metrics are in CloudWatch, they can be graphed together to observe their relationship. The incorrect options fail for various reasons: Using CloudWatch basic monitoring (first and fourth options) will only collect NetworkIn at 5-minute intervals, not the required 60 seconds. CloudWatch does not natively collect MemoryBytesUsed (or memused) for EC2 instances without the CloudWatch agent (first and second options).