AZ-400 Question 295
Select 2Your DevOps team is investigating intermittent performance degradation in a production web application running on multiple Azure Virtual Machines. The application experiences occasional CPU spikes, high memory consumption, and fluctuating disk I/O and network throughput. You suspect these resource bottlenecks are causing the slowdowns. Which two approaches can help you effectively inspect and diagnose CPU, memory, disk, and network performance indicators in this environment?
- A
Deploy and configure Azure Monitor or the Log Analytics agent on each VM to collect and analyze performance counters for CPU, memory, disk, and network usage.
- B
Enable ephemeral OS disks on each VM to reduce operational overhead and automatically expose granular disk performance metrics.
- C
Use the Azure Monitor Metrics blade to view real-time and historical charts for CPU, memory, disk, and network usage across the VMs.
- D
Rely solely on application logs to interpret and troubleshoot resource utilization across the entire stack.
Show answer and explanation
Correct answers: A, C
Explanation
To diagnose infrastructure-level performance bottlenecks, it is crucial to gather and visualize low-level metrics like CPU usage, memory consumption, disk I/O, and network performance. Deploying the Azure Monitor or Log Analytics agent on each VM ensures that these metrics are collected in detail. The Azure Monitor Metrics blade then allows you to accurately track, correlate, and analyze these metrics over time, helping to pinpoint and resolve performance issues. For more information, consult Microsoft documentation on Azure Monitor and Log Analytics agents (https://learn.microsoft.com/azure/azure-monitor/).
- A. Correct.
Correct. By installing and configuring Azure Monitor or the Log Analytics agent on each VM, you can capture detailed performance counters for CPU, memory, disk, and network usage. This agent-based approach allows you to centralize data for more granular analysis.
- B. Incorrect.
Incorrect. While ephemeral OS disks can improve certain I/O scenarios, they do not inherently provide deeper insights into CPU, memory, disk, or network metrics. Enabling ephemeral disks is primarily a storage option rather than a performance monitoring strategy.
- C. Correct.
Correct. Azure Monitor provides built-in metric views for CPU, memory, disk I/O, and network usage. Through the Metrics blade, you can analyze real-time and historical data, correlate trends, and set up alerts for performance anomalies.
- D. Incorrect.
Incorrect. Although application logs can help in understanding the behavior of the application itself, they typically do not offer comprehensive or reliable visibility into infrastructure-level performance indicators such as CPU, memory usage, disk I/O, and network throughput.