Multi-zonal and regional clusters

Resiliency is one of the primary concerns when deploying and operating enterprise-scale workloads. Having the ability to deploy workloads to multiple zones and regions is a big part of what makes resiliency possible. For instance, you might deploy some workloads to Us-central1-a and Us-central1-b, as shown in the following diagram:

GKE currently allows for clusters to be deployed to either a single zone, multiple zones, or to a region. Deploying a cluster to a region translates to deploying that cluster to every zone within that region.

Much akin to the supporting Kubernetes system, GKE is designed to orchestrate highly scalable enterprise container workloads. Reliability, scalability, and manageability are core value propositions for GKE. Applications, services, and other workloads that fit this bill will be a good fit for GKE. You can imagine a direct correlation between a workload being a fit for GKE and that workload's need to be scalable and resilient.