Build a failover strategy inside GCP to redirect if compute engine stop responding or is on an error state

Question

Currently I have a architecture inside GCP (Google Cloud Plataform) of a Wordpress application that works this way

Cloud DNS -> Load Balancer (with Cloud CDN enabled) -> Compute Engine (with wordpress application inside)

All components is inside GCP. I need to redirect to a specific URL when someone visit my wordpress application when Wordpress application becomes unresponsive or return errors. With another words, I need to do a failover strategy to a external url.

How this can be setup using GCP resources?

Answer 1

You may want to consider migrating your instance to MIG(Managed Instance Group). From there, the IP Address can be preserved to avoid having IP issues when doing failover from one instance to another if one fails. As reference, you can check this documentation on how MIG works and migrating your existing workload to a stateful MIG.