Question

A retail company runs a stateless web application on Compute Engine instances behind a global external Application Load Balancer. The application currently serves users from a single region (us-central1) using a zonal managed instance group. Leadership requires that the application survive the failure of an entire zone with no manual intervention and minimal capacity loss, while keeping the design as simple and cost-effective as possible. Which change to the compute and load-balancing design best meets this resilience requirement?

Accepted Answer

Convert the zonal managed instance group into a regional managed instance group spread across at least three zones in us-central1, keeping it as a backend of the global Application Load Balancer.

Answer

Deploy identical zonal managed instance groups in two additional regions and rely on the global Application Load Balancer to fail over between regions when a zone fails.

Answer

Keep the zonal managed instance group but enable autohealing health checks so failed instances are recreated automatically within the same zone.

Answer

Replace the global Application Load Balancer with a regional internal passthrough Network Load Balancer and add a second zonal instance group as a failover backend.