论文部分内容阅读
负载均衡在集群系统研究领域是一个比较重要的问题,它的八种调度算法是任务分配的核心,详细介绍了LVS中的调度理论,研究并讨论了常用算法的不足,在分析这些算法优缺点的基础上,提出了一种基于反馈的负载均衡算法理论,该算法引入一个负载容余参数可以更准确地描述集群节点的负载状态,在考虑各个服务节点处理能力真实负载的基础上,尽量简化负载均衡器的调度算法。经测试,其结果显示改进后的算法优于改进前的算法。
Load balancing is a more important issue in the field of cluster system research. Its eight scheduling algorithms are the core of task assignment. The scheduling theory in LVS is introduced in detail. The deficiencies of commonly used algorithms are studied and discussed. The advantages and disadvantages of these algorithms , This paper proposes a load-balancing algorithm based on feedback theory. This algorithm introduces a load-tolerance parameter to describe the load status of cluster nodes more accurately. On the basis of considering the actual load of each service node, Load balancer scheduling algorithm. Tested, the results show that the improved algorithm is better than the improved algorithm.