A Phased Scheduling Method In Workflow And Scheduling Trigger Mechanism

In the modern society with the high-speed development of information,workflow technologybecomes more and more important in the fields of computer.In the complicated system,theworkflow system deal with every task by scheduling the corresponding resources, so that theefficiency of the whole system is improved. The workflow scheduling algorithm is an importantpart of the whole workflow system.The workflow scheduling algorithm is divided into static scheduling algorithm,dynamicscheduling algorithm and phased scheduling algorithm. Existing phased scheduling algorithmcombines the advantages of both before, but when the task execution time is longer, there aresome defects about the scheduling trigger mechanism of this algorithm,so the efficiency of thetask concurrency will be affected.This thesis proposes an improved phased scheduling algorithm.The aim of this algorithm isto shorten the overall execution time of the workflow system in the environment of longerexecution time. This thesis adopts the genetic algorithm to optimize the configuration of tasks andresources, so that in the case of the known quantity and execution time of resources, adoptingappropriate chromosome coding and appropriate genetic manipulation to get possibly superiorcombination of tasks and resources configuration and shorten the overall task completion time.This thesis put forward two improvements on the original scheduling trigger mechanism ofunimproved algorithm. First, add a new scheduling trigger point. Second,add a variable timewindow to promote instance concurrency execution more effective.Experimental results show thatthe improved algorithm improves the efficiency of concurrent execution among the tasks of theworkflow system and shorten the overall execution time of the workflow system in a certainextent.