Your company's critical application is hosted on Compute Engine managed instance groups. To ensure the instances are resilient and highly available without manual intervention, how should the managed instance group be configured?