You have a production application deployed on Compute Engine VM instances managed by an instance group, with traffic routed through an HTTP(s) load balancer. Users are unable to access the application. What monitoring technique should you implement to receive alerts when the application becomes unavailable?