AZ-400 Question 16
Select 2Your organization deploys microservices to Azure Kubernetes Service (AKS) using Azure DevOps pipelines. The team wants to measure how quickly changes move from commit to production (lead time), how often deployments occur, and how to rapidly detect performance bottlenecks. Which two of the following approaches would best help you design and implement appropriate metrics and queries for these requirements?
- A
Collect and analyze DORA metrics (including lead time and deployment frequency) using Azure DevOps Pipeline telemetry and Azure Monitor custom queries
- B
Rely solely on manual spreadsheets to record snapshot data from each deployment pipeline run
- C
Create a custom Azure Monitor metric to track build success across pipelines, storing pass/fail status as time series data for trending analysis
- D
Retain logs only within containerized workloads on AKS for faster performance, without aggregating into a centralized system like Azure Monitor
Show answer and explanation
Correct answers: A, C
Explanation
When designing DevOps metrics and queries, it is critical to capture industry-standard metrics (like DORA) and store them in a real-time, centralized system such as Azure Monitor or Application Insights. Azure Monitor queries allow teams to correlate data (e.g., lead time, deployment frequency), identify performance outliers, and rapidly respond to issues. Refer to official Microsoft DevOps documentation on measuring DevOps performance (https://docs.microsoft.com/azure/devops) and the DORA research program for guidance on implementing these practices effectively.
- A. Correct.
Collecting DORA metrics (like lead time for changes and deployment frequency) through Azure DevOps telemetry and analyzing them with Azure Monitor queries is an industry best practice. It provides real-time insights into delivery performance, making it easier to identify bottlenecks. This is a correct approach.
- B. Incorrect.
Manual spreadsheets are not as reliable or timely for tracking DevOps metrics. They increase the chance of human error, lack real-time updates, and do not scale well. This option does not align with recommended practices for designing automated metrics and queries.
- C. Correct.
Creating a custom Azure Monitor metric for build success rates is beneficial for understanding continuous integration performance over time. By storing results as time-series data, teams can detect trends and quickly investigate failures. This is a valid approach.
- D. Incorrect.
Relying solely on ephemeral logs kept within the container cluster does not provide long-term visibility or the ability to conduct cross-service queries. Without aggregation into a centralized solution like Azure Monitor or Log Analytics, teams cannot effectively track or diagnose issues. This is not recommended.