Create Compute Engine resources in us-central1-b. Balance the load across both us-central1-a and us-central1-b.

Create a Managed Instance Group and specify us-central1-a as the zone. Configure the Health Check with a short Health Interval.

Create an HTTP(S) Load Balancer. Create one or more global forwarding rules to direct traffic to your VMs.

Perform regular backups of your application. Create a Cloud Monitoring Alert and be notified if your application becomes unavailable. Restore from backups when notified.