Fuzzy-Based Resource Reallocation Scheduling Model in Cloud Computing

摘要

A cloud computing system consists of physical resources for processing large-scale tasks. With a recent trend of rapidly growing data, a cloud computing system needs a processing method to process a large-scale task in a physical resource. Generally, a physical resource divides a requested large-scale task to several tasks. And a processing time of each divided task varies with two factors which are processing efficiency of each resource and distance between resources. Although a resource completes a task, the resource is standing by until all divided tasks are completed. When all resources complete a large-scale task, each resource can start to process a next task. In this paper, we propose a Fuzzy-based Resource Reallocation Scheduling Model (FRRSM). Using fuzzy rule, FRRSM reallocates an uncompleted task to with a resource in considering efficiency and distance factors of the resource. FRRSM is an efficient method for processing a large-scale task or multiple large-scale tasks.