Your web app runs on multiple instances in Azure App Service. Occasionally, one instance becomes unresponsive. You want the platform to automatically detect an unhealthy instance and remove it from the load balancer, then replace it without manual intervention. What should you do?

The Health check feature in Azure App Service is designed for this scenario. Enabling it and specifying an endpoint (e.g., /healthz) allows the App Service platform to periodically ping that endpoint. If the endpoint consistently returns an HTTP status code outside of the 200-299 range, or fails to respond, the instance is deemed unhealthy and automatically removed from the load balancer. A new instance is then provisioned to replace it, ensuring high availability without manual intervention. The WEBSITEHEALTHCHECKMAXPINGFAILURES app setting controls how many consecutive failed health probes trigger an instance replacement. Lowering this value makes the detection and replacement process faster, which is beneficial for quickly addressing unresponsive instances. Incorrect options: Application Insights availability tests monitor external reachability but don't directly integrate with App Service's instance management for automatic replacement. Always On keeps the app loaded but doesn't provide health-based instance replacement. Traffic Manager is a DNS-based load balancer and doesn't manage individual App Service instance health within a single App Service Plan.