MLA-C01 Question 403
Select 3You are a Machine Learning Engineer at a company using AWS Lambda functions to preprocess data for a real-time machine learning pipeline. Recently, the pipeline has been experiencing increased latency, and you suspect the issue lies within the Lambda function execution. Which combination of tools would you use to identify and troubleshoot the root cause of the latency issue?
- A
AWS X-Ray to analyze the end-to-end trace of the Lambda function and identify where latency is occurring
- B
Amazon CloudWatch Metrics to monitor the duration and error rates of the Lambda function
- C
Amazon CloudWatch Logs Insights to query and analyze detailed logs generated by the Lambda function
- D
AWS Step Functions to directly debug Lambda function execution
- E
AWS Trusted Advisor to check for service limits that might be causing the Lambda function to throttle
Show answer and explanation
Correct answers: A, B, C
Explanation
To troubleshoot latency in Lambda functions, a combination of AWS X-Ray for tracing, Amazon CloudWatch Metrics for monitoring performance, and Amazon CloudWatch Logs Insights for log-level analysis is most effective. These tools together provide a comprehensive view of the function's behavior, enabling you to pinpoint the root cause of latency in the pipeline.
- A. Correct.
AWS X-Ray provides end-to-end tracing, which is critical for identifying where latency is occurring in a distributed system, including within Lambda functions.
- B. Correct.
Amazon CloudWatch Metrics offers metrics like duration, error count, and invocation frequency for Lambda functions, which can help identify performance bottlenecks.
- C. Correct.
Amazon CloudWatch Logs Insights allows you to query and analyze logs from Lambda functions to identify specific errors or performance issues in the code execution.
- D. Incorrect.
AWS Step Functions is useful for orchestrating workflows but is not a direct debugging or monitoring tool for Lambda functions.
- E. Incorrect.
AWS Trusted Advisor offers recommendations on cost optimization, security, and service limits, but it is not designed for real-time latency troubleshooting.