Question

A CloudOps engineer supports a microservices application running on Amazon ECS behind an Application Load Balancer. Users report intermittent slow response times, but CloudWatch metrics for CPU, memory, and ALB target response time all appear within normal ranges. The team needs to identify which specific downstream service or dependency (such as a database call or third-party API) is introducing the latency for affected requests. Which approach best provides this insight?

Accepted Answer

Instrument the services with the AWS X-Ray SDK and use the X-Ray service map and trace analysis to identify segments and subsegments with high latency

Answer

Enable CloudWatch Container Insights and create a dashboard showing aggregated CPU and memory utilization per ECS task

Answer

Create a CloudWatch metric filter on the ALB access logs to count HTTP 5xx responses and trigger an alarm

Answer

Configure CloudWatch anomaly detection on the ALB RequestCount metric to detect unusual traffic patterns