基于QoS的网格负载调度算法  

Grid Load Schedule Algorithm Based on QoS

作　　者：雷炳翰[1] 何军[1] 何翔[1] 李奇 

机构地区：[1]四川大学计算机学院,成都610065 [2]四川省计算机软件重点实验室,成都610068

出　　处：《计算机工程》2009年第24期96-98,共3页Computer Engineering

基　　金：国家"863"计划基金资助项目(2008AA01Z208);国家部委基金资助项目;四川省科技厅基金资助项目(2007H12-010);四川省计算机软件重点实验室基金资助项目(08JSU002)

摘　　要：针对多数启发式算法不能满足任务紧迫性需求的问题,基于Min-min算法和Sufferage算法的优点提出基于紧迫性需求的系统损失与任务复杂度平衡(QoS-based BSLTC)算法。在网格任务中加入QoS参数,在任务调度过程中,基于任务的分配损失和权衡系数,比较最大损失任务和最高计算复杂度任务,选出对系统性能影响最大的任务,并将该任务映射到完成时间最早的资源。分析结果表明,与Min-min算法和Sufferage算法相比,QoS-based BSLTC算法具有执行效率高、任务完成总时间短、负载均衡度高、兼顾任务紧迫性的优点。Aiming at the problem that most of the heuristic algorithm can not meet the task’s urgency demand,this paper presents the QoS-based Balance System Loss and Task Complexity（QoS-based BSLTC） algorithm based on the merits of Min-min algorithm and Sufferage algorithm.It joins the QoS parameter in the grid task.During the process of task schedule,the maximal loss task and the maximal computational complexity task is compared based on the distribution loss and measure coefficient.The task which has the most influence on the system is picked out and mapped to the resource with a minimum completion time.Analysis results verify that QoS-based BSLTC algorithm has merits of higher executing efficiency,needing less time to complete the task,high load balance degree and giving consideration to task urgency by comparing with Min-min algorithm and Sufferage algorithm.

关 键 词：任务调度 QOS参数 负载均衡度 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]