The company needs to monitor its public website and send an Amazon SNS notification if availability falls below 99% as experienced by end users. Which design will produce an accurate user-experience view and meet the requirement?

The correct answer is to use a CloudWatch Synthetics heartbeat canary. A heartbeat canary actively simulates user traffic to your website, making requests and monitoring the responses. This provides a direct measure of user experience and availability. The SuccessPercent metric from this canary accurately reflects the percentage of successful requests, allowing you to set an alarm for availability falling below 99% and trigger an SNS notification. Incorrect options: Monitoring HTTP 4xx and 5xx counts from web server logs only indicates server-side errors, not necessarily end-user experience, as network issues or client-side problems could still affect availability. Using anomaly detection on published CloudWatch metrics might detect unusual patterns, but it doesn't directly measure user-experienced availability against a specific percentage threshold like 99%. A broken link checker canary focuses on finding broken links within the site, not on overall site availability or user experience in the same way a heartbeat canary does.