AZ-400 Question 18
Single answerYou manage an Azure DevOps environment with multiple CI/CD pipelines for various microservices. Senior leadership wants to see a visual dashboard showing the average duration of each pipeline stage, as well as trends over time to detect any performance degradations. You decide to store and query these metrics in Azure Monitor using Kusto (KQL) queries for more advanced analytics and dashboards. Which solution should you implement to meet these requirements?
- A
Use Azure DevOps Analytics views alone for real-time dashboards on stage duration without sending data to Azure Monitor.
- B
Add a step in each pipeline to send stage timing data to Azure Log Analytics and build Kusto-based dashboards in Azure Monitor.
- C
Rely on the default pipeline run logs in Azure DevOps to automatically feed stage metrics to Azure Monitor without additional configuration.
- D
Set up a self-hosted agent metric collector that writes raw pipeline logs to local storage and manually import them into Azure Monitor.
Show answer and explanation
Correct answer: B
Explanation
To design and implement appropriate metrics and queries for DevOps, you need to gather data about pipeline stages and store that information in a service capable of advanced querying and visualization. By adding a custom step in each pipeline to publish timing data to Azure Log Analytics, you can leverage Azure Monitor� Kusto (KQL) queries for trend analysis and more granular insights. This approach aligns with Microsoft best practices, where Azure Log Analytics serves as a central repository to build real-time dashboards, investigate anomalies, and maintain historical pipelines performance data. See Microsoft documentation on 'View and query pipeline logs in Azure Monitor' and 'Azure DevOps Analytics' for further guidance.
- A. Incorrect.
Option 1 is incorrect. While Azure DevOps Analytics provides useful insights, it does not offer the same level of advanced Kusto-based querying and visualization in Azure Monitor. Using Analytics views alone would not directly meet the need for advanced custom dashboards in Azure Monitor.
- B. Correct.
Option 2 is correct. Incorporating a step in each pipeline to publish timing data (e.g., via logs or custom telemetry) to Azure Log Analytics allows you to ingest that data into Azure Monitor, where you can create advanced dashboards and Kusto queries to visualize and trend the metrics over time.
- C. Incorrect.
Option 3 is incorrect. Although Azure DevOps does create logs, simply relying on default run logs does not automatically export metrics to Azure Monitor. Additional configuration or pipeline steps are needed to send data to Log Analytics workspace for Kusto-based analytics.
- D. Incorrect.
Option 4 is incorrect. Writing logs to local storage and then manually importing them into Azure Monitor is cumbersome, prone to errors, and not scalable. It also does not address the need for near real-time metrics and automated reporting.