AI-102 Question 64
Select 3You are managing an Azure AI resource deployed for a machine learning solution. Users have reported slower response times and occasional downtime. You want to monitor the resource to identify performance issues and ensure reliability. Which actions should you take to effectively monitor the Azure AI resource?
- A
Enable Azure Monitor Metrics to track resource performance such as CPU, memory, and latency.
- B
Configure Application Insights to monitor request rates, failures, and response times for the AI resource.
- C
Set up an Activity Log Alert to track changes to the resource configuration.
- D
Enable a Load Test in Azure DevOps to simulate user traffic and monitor the AI resource's performance.
- E
Use Azure Advisor to track the AI resource's cost optimization recommendations.
Show answer and explanation
Correct answers: A, B, C
Explanation
To effectively monitor an Azure AI resource, you need to leverage Azure Monitor Metrics for performance tracking, Application Insights for request and failure insights, and Activity Log Alerts for configuration change tracking. These tools together provide comprehensive monitoring and diagnostics for ensuring the reliability and performance of the resource.
- A. Correct.
Correct. Azure Monitor Metrics allows you to gather detailed performance data, such as CPU usage, memory, and latency, which are essential for tracking the health of the AI resource.
- B. Correct.
Correct. Application Insights is crucial for monitoring request rates, failures, and response times, helping you identify and resolve performance bottlenecks in the AI resource.
- C. Correct.
Correct. Activity Log Alerts are important for tracking configuration changes, which might be contributing to performance issues or downtime.
- D. Incorrect.
Incorrect. While Load Testing can simulate traffic, it is not directly part of monitoring the AI resource in Azure. It is more relevant to testing the application's scalability rather than real-time monitoring.
- E. Incorrect.
Incorrect. Azure Advisor provides cost and performance recommendations but does not directly monitor the AI resource's health or performance metrics.