Cloud Resource Reclamation Sequencing for Dependency-Safe Shutdowns

The patent segments the cloud computing resources into hierarchical levels (first level, second level, etc.) based on their dependency relationships. Each level can be managed independently during shutdown operations, allowing the system to handle distributed multi-region deployments without requiring global coordination of all resources simultaneously. This segmentation reduces the complexity of determining dependencies across distributed regions.

The patent implements preliminary actions by first identifying and shutting down dependent resources (second level) before shutting down the primary resources (first level). This preliminary ordering of shutdown actions ensures that data dependencies are respected, preventing data loss and corruption, while also optimizing the overall shutdown duration by systematically progressing through the resource hierarchy.

The patent changes the state parameter of cloud computing resources from a binary state (active/inactive) to a multi-state model that includes an intermediate suspended state. Resources can be transitioned to this suspended state during reclamation, allowing them to be temporarily removed from active use to free up hardware while maintaining the capability to recover and restart them later if needed, thus balancing resource utilization with recoverability.