An administrator is managing an Oracle Container Engine for Kubernetes (OKE) cluster. They want to ensure that worker nodes are distributed across multiple failure domains or Availability Domains to prevent downtime if a single hardware rack fails. How should this be configured?