一种新型的云任务调度算法研究

Research of a novel cloud task scheduling algorithm

云计算具有弹性、保证服务质量和按需的资源配置模型等特征,通常用于处理大批量的计算任务,因此任务调度策略对资源使用效率起着至关重要的作用.考虑到任务的数量和到达服务器的时间不确定性,并且用户对任务的执行往往有一定的期望(如任务优先级、执行时间等),如何合理地分配计算资源,最大程度满足用户的服务质量需求是一个值得研究的问题.为此,提出了一种新型的云环境下QoS-aware服务质量感知的任务调度算法(QTS),该算法结合贪心算法的思想,并加入了任务完成满意度模型作为任务调度的评价依据.通过扩展CloudSim仿真平台进行实验,将QTS与RR调度、Max-Min和Min-Min调度比较,结果表明,QTS是一种有效的任务调度算法.

With its flexibility , guaranteed quality of service and on-demand features such as resource allocation model , cloud computing is often used to handle large computing tasks , so efficient task scheduling strategies of cloud computing play a vital role .Given the uncertainty in the number of tasks and the time of arrival at the server ,and the fact that ,users tend to have certain expectations （such as task priority ,execution time ,etc .） for the implementation of the tasks ,reasonable allocation of computing resources for task scheduling to satisfy users＇QoS requirements is of great importance .A novel QoS-aware task scheduling mechanism （QTS） was proposed ,this scheduling mechanism can best meet the user＇s QoS requirements .By comparing QTS with RR , Max-Min and Min-Min scheduling policies by CloudSim simulation ,it was found that QTS is a more effective task scheduling mechanism .