Your company has developed an application intended to be deployed on Google Kubernetes Engine (GKE). This application comprises multiple components, where some components are non-fault-tolerant and can tolerate downtime, while other components are critical and require constant availability without any downtime. Given these requirements, you need to configure a GKE cluster in a way that ensures high availability for the critical components and minimizes operational costs for the overall deployment. How should you configure the GKE cluster to achieve these goals?