基于负载均衡的任务调度算法

基于传统任务调度算法, 通过在调度过程中增设服务器最佳期望序列及负载均衡指数, 采用任务连接数约束服务器选择, 解决了传统任务调度算法对于集群负载均衡性能的忽略问题. 实验结果表明, 该改进算法能达到较高负载均衡度, 同时缩短任务完成时间, 并提高了集群对于批量级任务的处理能力, 从而达到了优化的目的.

Algorithm for Tasks Scheduling Based on Load Balance

On the basis of traditional tasks scheduling algorithm, we established the optimal sequence of servers and load balance index for tasks in the process of scheduling. Meanwhile, we used a number of connections to realize the constraint of servers, which has improved the performance of load balancing that has ignored in the traditional task scheduling algorithm. The experimental results show that our improved algorithm can achieve the high degree of load balancing and shorten the task completion time. In addition, this algorithm improves the ability of batching task processing for server cluster, thereby achieving the goal of optimization.