A managed instance group serves an HTTP application on port 8080. Which health check should you configure on the load balancer backend to accurately determine instance health?

The correct choice is an HTTP health check probing /health on port 8080 because it directly validates the application's availability and responsiveness on its serving port. This ensures the load balancer only directs traffic to instances that can successfully handle HTTP requests on the specified port and path. A TCP health check on port 80 would only confirm the port is open, not that the application is functional, and it's on the wrong port. An HTTPS health check on port 443 with a different domain is incorrect because the application serves HTTP on port 8080, not HTTPS on 443, and a different domain is irrelevant. Finally, health checks are crucial for load balancers to prevent sending traffic to unhealthy instances, so "no health check" is incorrect.