Question

A developer is troubleshooting intermittent high latency in a serverless order-processing API. The API Gateway invokes a Lambda function that calls DynamoDB and an external payment provider over HTTPS. The team has enabled AWS X-Ray active tracing on the API and Lambda. When they open the X-Ray service map, they need to quickly determine which downstream dependency is responsible for the added latency during slow requests. What is the MOST effective way to isolate the failing component using X-Ray?

Accepted Answer

Examine the service map and trace details to compare the duration of the DynamoDB and payment-provider subsegments within the Lambda function's segment.

Answer

Enable CloudWatch Lambda Insights and read the aggregated CPU and memory metrics to infer which call is slow.

Answer

Increase the Lambda function memory allocation and re-run the requests to see if overall latency decreases.

Answer

Add a CloudWatch alarm on the API Gateway Latency metric and wait for it to breach before investigating logs.