Monitor and Troubleshoot Azure Solutions AI-200 Practice Questions (Page 6)

Questions 26–30

1. 26expert · hard

   A company has a microservices application with OpenTelemetry tracing enabled. They notice that a specific trace shows a 'CheckoutService' span with a duration of 2 seconds, but the child spans (PaymentService and InventoryService) each show durations of 200 ms. The trace also shows a 'Database' span with a duration of 1.5 seconds. What is the most likely bottleneck?

   • AThe CheckoutService itself is the bottleneck because it has the longest duration.
   • BThe InventoryService is the bottleneck because it is a child of CheckoutService.
   • CThe PaymentService is the bottleneck because it is a child of CheckoutService.
   • DThe database query is the bottleneck because it accounts for most of the CheckoutService span duration.
   Reveal Solution Discussion Select an answer first
2. 27expert · hard

   A company has a distributed system with OpenTelemetry tracing. They notice that a trace for a user request shows a 'PaymentService' span with a duration of 1 second, but the 'PaymentService' span has no child spans. The trace also shows that the 'OrderService' span (parent) has a duration of 1.5 seconds. What is the most likely cause of the missing 0.5 seconds?

   • AThe PaymentService is using a queue to process payments asynchronously.
   • BThe PaymentService is making a call to an external payment gateway that is not instrumented with OpenTelemetry.
   • CThe PaymentService span is using a different trace ID than the OrderService span.
   • DThe sampling rate is set to 10%, so some spans are missing.
   Reveal Solution Discussion Select an answer first
4. 28application · medium

   A software company is building a new microservices application on Azure and wants to implement distributed tracing from the start. The team needs to choose a technology that provides a vendor-neutral way to generate traces, supports multiple programming languages, and can export data to Azure Monitor. Which approach should the team take?

   • AUse the Application Insights SDK directly in each service and configure the instrumentation key to send data to Application Insights.
   • BUse Azure Monitor's built-in auto-instrumentation for all services without adding any SDKs.
   • CUse the OpenTelemetry SDKs in each service and configure the Azure Monitor exporter to send trace data to Application Insights.
   • DUse a custom logging library that writes trace IDs to the application logs and then query them with KQL.
   Reveal Solution Discussion Select an answer first
5. 29foundation · easy

   What is the basic structure of a KQL query?

   • AA data source followed by a series of operators separated by pipes (|)
   • BA single function call that returns a result set
   • CSELECT column FROM table WHERE condition
   • DA series of commands enclosed in curly braces
   Reveal Solution Discussion Select an answer first
6. 30application · medium

   A support engineer is troubleshooting a web app that intermittently returns HTTP 500 errors. They want to find the correlation between error responses and high CPU usage. The application logs are in 'AppLogs' and CPU metrics are in 'InsightsMetrics'. What should they do?

   • ALook at the AppLogs table alone to find the error messages.
   • BCreate a separate workbook for logs and another for metrics.
   • CWrite a KQL query that joins AppLogs and InsightsMetrics on a common timestamp field and look for patterns.
   • DLook at the InsightsMetrics table alone to find CPU spikes.
   Reveal Solution Discussion Select an answer first