AZ-900 Describe Azure architecture and services Practice Question

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✗

  Azure Traffic Manager health probes

  Why it's wrong here

  Azure Traffic Manager operates at the DNS layer and uses health probes only to assess whether an endpoint (like a cloud service or web app) is alive and can receive traffic, so it can route users to the next available endpoint during failover. These probes check basic TCP/HTTP connectivity from Traffic Manager's edge infrastructure, but they do not measure end-user response times or compare availability across distant geographic regions. Application Insights Availability tests instead issue actual HTTPS requests from globally distributed Azure points of presence to validate the app's availability from the user's perspective. Thus, Traffic Manager health probes are a routing mechanism, not a global availability monitoring solution.

• ✓

  Application Insights Availability tests

  Why this is correct

  Application Insights Availability tests are the correct tool because they actively send synthetic web requests to your application from multiple global locations, checking whether the endpoint is reachable and how quickly it responds. You can configure host URLs and expected HTTP status codes, and the tests send telemetry such as 'availability percentage' and 'response times' into Azure Monitor. The tests run on a schedule from Azure's worldwide data centers, giving you an outside-in view of availability and latency that reflects what actual users experience. This global monitoring capability is exactly what the scenario requires, unlike routing-oriented probes or notification-only action groups.

• ✗

  Azure Monitor Action Groups

  Why it's wrong here

  Azure Monitor Action Groups are purely a response mechanism: they define channels such as email, SMS, webhook, Microsoft Teams, or ITSM to notify operators when an alert rule is triggered, and they can also run an automation runbook. They do not themselves generate any availability signal or measure an application's state; they only apply the configured notification actions after a signal arrives from a resource such as Application Insights. Even if an action group is attached to a synthetic alert rule, the actual monitoring is performed by the availability test, not by the action group. Therefore, action groups are a downstream alerting component, not the tool that performs global availability monitoring.

• ✗

  Azure Load Balancer health probes

  Why it's wrong here

  Azure Load Balancer health probes check the health of virtual machine instances in the backend pool by sending periodic TCP or HTTP requests to a defined port and path, and then the load balancer only forwards new connections to healthy instances. These probes are scoped to the load balancer's virtual network and backend set; they never originate from the internet or from multiple geographic locations, so they cannot reflect how your application performs for users around the world. Their sole purpose is to aid in traffic distribution and failover at the infrastructure layer, not to measure user-facing availability. Application Insights Availability tests provide the required global, end-user-focused monitoring, making Load Balancer probes an incorrect choice.