Cloud Storage Scheduling Using Time-Series Topology for Load Balancing

The patent extends the scheduling dimension from single-capacity dimension to multi-dimensional space by incorporating performance metrics (IOPS, bandwidth, latency) alongside capacity. This dimensional expansion enables the system to make more informed scheduling decisions that balance both capacity and performance, thereby improving response speed while maintaining manageable complexity through structured multi-criteria evaluation.

The patent introduces multiple scheduling parameters beyond traditional capacity metrics, including performance parameters such as IOPS, bandwidth, and latency. By changing the parameter set from single-dimension to multi-dimensional, the system achieves more uniform resource allocation that simultaneously optimizes for capacity utilization and response speed, resolving the contradiction between simplicity and productivity.

The patent segments the scheduling process into distinct modules: capacity evaluation module, performance evaluation module, and scheduling decision module. This segmentation allows each dimension (capacity and performance) to be evaluated independently through dedicated functions, then combined systematically. The modular approach improves load balancing efficiency while managing complexity by breaking down the multi-dimensional scheduling problem into manageable components.

The patent implements a feedback mechanism where time-series statistical data about resource states is continuously collected and fed back into the scheduling system. This feedback loop enables the system to adjust scheduling decisions based on actual performance observations, improving load balancing efficiency while keeping complexity manageable through automated adaptive control rather than complex manual configuration.