SCS-C02 Question 48
Single answerA security operations team is using Amazon CloudWatch to monitor an application running on AWS. They want to identify anomalies in their application’s behavior by using visualizations that can detect outliers in metrics such as request latency and error rates. Which approach should the team take to implement anomaly detection effectively?
- A
Enable CloudWatch Anomaly Detection on specific metrics and use the anomaly detection band in dashboards to visualize outliers.
- B
Create custom metrics in CloudWatch and manually analyze data trends to detect anomalies.
- C
Enable AWS Config rules to monitor and visualize changes in metrics over time.
- D
Use AWS Trusted Advisor to generate visualizations for identifying anomalies in application performance.
Show answer and explanation
Correct answer: A
Explanation
The best approach to identifying anomalies in application metrics is to use Amazon CloudWatch Anomaly Detection. This feature integrates machine learning to establish a baseline for normal metric behavior and provides an anomaly detection band for visualizing outliers. It is specifically designed for such use cases, making it a more efficient and reliable solution than manual analysis or tools not intended for metric monitoring.
- A. Correct.
Correct: CloudWatch Anomaly Detection allows you to apply machine learning models to metrics and visualize anomalies using an anomaly detection band in dashboards, making it an effective way to identify outliers.
- B. Incorrect.
Incorrect: While you can create custom metrics, manually analyzing trends is time-consuming and error-prone compared to using automated tools like CloudWatch Anomaly Detection.
- C. Incorrect.
Incorrect: AWS Config is primarily used for resource configuration compliance and does not visualize metric trends or detect anomalies.
- D. Incorrect.
Incorrect: AWS Trusted Advisor provides recommendations for cost optimization, security, and performance improvements but does not generate metric visualizations for anomaly detection.