AWS Certified Solutions Architect Professional SAP-C02 Practice Question
A global travel-booking company runs its REST booking service on Amazon EKS behind an Application Load Balancer (ALB). Management has introduced a new performance objective: "95 % of booking requests must complete within 200 ms." The operations team already receives average CPUUtilization and request-count alarms from Amazon CloudWatch, but the new objective has not been instrumented. You must implement an alert that notifies the team whenever the service falls out of compliance while requiring least development effort.
Which approach most directly translates the business requirement into measurable metrics and alarms?
Enable AWS X-Ray tracing at a 100 % sampling rate for the EKS cluster and configure an alarm on the X-Ray ServiceAverageLatency metric when average latency exceeds 0.2 s.
Instrument the booking service to publish a custom BookingLatency metric via the CloudWatch Embedded Metric Format, then build a metric-math expression that calculates p95 and alarms when it exceeds 0.2 s.
Create a CloudWatch alarm on the ALB TargetResponseTime metric using the p95 statistic and a 0.2-second threshold for the booking target group.
Create a target-tracking Auto Scaling policy that scales the number of ALB-attached pods when ALB RequestCountPerTarget surpasses 1 000 requests per minute to keep latency below 200 ms.
TargetResponseTime is a built-in Application Load Balancer metric that records the time (in seconds) from when the request leaves the load balancer until the target starts to send the response headers. Because CloudWatch supports percentile statistics, an alarm on the p95 value of this metric with a 0.2-second threshold expresses the requirement that 95 % of requests finish within 200 ms. No application changes or additional infrastructure are required, so this solution meets the "least development effort" criterion.
Publishing a custom BookingLatency metric would also work, but it demands code changes, metric maintenance, and permissions. AWS X-Ray only provides average latency by default and does not expose the 95th percentile; enabling 100 % sampling can introduce overhead. An Auto Scaling policy that reacts to request count may reduce latency indirectly, but it does not measure or alarm on latency and therefore cannot prove the SLA is met.
Ask Bash
Bash is our AI bot, trained to help you pass your exam. AI Generated Content may display inaccurate information, always double-check anything important.
What is the TargetResponseTime metric in CloudWatch?
Open an interactive chat with Bash
What is the p95 statistic in CloudWatch, and why is it important?
Open an interactive chat with Bash
Why is using a custom BookingLatency metric less efficient for this use case?
Open an interactive chat with Bash
AWS Certified Solutions Architect Professional SAP-C02
Continuous Improvement for Existing Solutions
Your Score:
Report Issue
Bash, the Crucial Exams Chat Bot
AI Bot
Loading...
Loading...
Loading...
IT & Cybersecurity Package Join Premium for Full Access