【摘要】 用户任务的随机到达以及用户对任务的QoS需求,不但要求异构分布式系统满足用户QoS需求,而且要求系统尽可能多地处理随机到达的任务,并且使系统负载尽可能达到平衡,因而找到满足这些条件的调度策略具有重要意义。该文提出一种基于QoS Min-Min的元任务调度策略,使系统CPU得到充分利用,从而处理更多任务。和Min-Min调度算法相比,QoS Min-Min减少了任务淘汰率和提高了系统CPU利用率。模拟实验结果显示,该调度策略具有实效性。更多还原

【Abstract】 Since user task arriving at heterogeneous distributed system is random and user has requirement of service for the task,it needs to make heterogeneous distributed system satisfy user QoS,and make system transact tasks as much as possible,and it is an important target to make system keep load balance,so it is very important to find a good schedule strategy.A QoS guided meta-task schedule strategy is described,which can make distributed system use CPU sufficiently and can make system transact tasks as much as possible.Comparing to Min-Min schedule strategy,it reasonably reduces ratio of task discarded and improves CPU using ratio of the system.Experimental result shows it has actual effect.更多还原