Create an autoscaling managed instance group for an HTTPS web app and ensure unhealthy VMs are recreated. What should you do?

To ensure unhealthy VMs are recreated in an autoscaling managed instance group (MIG) for an HTTPS web app, you must configure a health check. The health check actively probes the instances to determine their health status. By specifying port 443, the health check verifies that the HTTPS service running on the instances is responsive. If an instance fails the health checks, the MIG considers it unhealthy and will attempt to recreate it, maintaining the desired service availability. Selecting Multi-Zone provides high availability across zones but doesn't inherently recreate unhealthy instances within a zone. Adding a label 'health-check' to the instance template is not how health checks are configured for MIGs. A startup script sending a heartbeat to the metadata server is a custom solution that doesn't integrate directly with the MIG's built-in health check and autohealing mechanisms.