You deployed a web app on Azure App Service (Basic plan, single region). Users report slow responses. You need to capture complete call stacks correlated across instances to diagnose performance, while minimizing cost and user impact. Which three actions should you take?

Enabling Application Insights site extensions is crucial for comprehensive monitoring. It instruments your application to collect telemetry data, including request rates, response times, and dependency calls, which are essential for diagnosing performance issues. Enabling Profiler allows you to capture detailed performance traces, including call stacks, which helps identify bottlenecks in your code. This is exactly what's needed to diagnose slow responses. Enabling the Always On setting prevents the app from being unloaded due to inactivity, ensuring it's always ready to respond to requests and eliminating cold start delays, which can be perceived as slow responses. Restarting apps is a temporary fix and doesn't provide diagnostic data. Upgrading the App Service plan might improve performance but doesn't diagnose the root cause and is not the most cost-effective initial step. Enabling Snapshot Debugger is for debugging exceptions, not general performance profiling.