摘要
随着各大数据中心的建立,数据中心达到成千上万台服务器的规模。计算节点之间存在差异性,传统调度算法很难满足计算中心资源调度,主机之间难免出现负载不均衡,直接影响了云平台提供服务的能力,造成资源浪费。根据数据中心的实际情况,基于现有环境下各种主流资源调度策略,针对资源分配与主机运行时动态调度两方面提出完整的资源调度策略方案,提高数据中心计算能力,避免不必要的资源浪费,显著改善数据中心服务质量,实现各种资源负载均衡。
With the establishment of the major data centers,the data center has grown into the scale of thousands of servers.There are all kinds of differences between compute nodes.Traditional scheduling algorithm can hardly satisfy the resource scheduling of computing center,the load imbalance of the usage between each host is inevitable,it will create a direct impact on the cloud platform′s ability to provide services and create a waste of resource.Aiming at allocation of resources and run-time dynamic scheduling,we propose a complete resource scheduling strategy based on various mainstream under the existing environment and data centers,which improves data center service quality and the computing ability of data center.It will also avoid unnecessary waste of resources,and finally realize load balance of various kinds of resources.
作者
张晓兵
朱仲马
ZHANG Xiao-bing;ZHU Zhong-ma(China Electronics Technology Group Corporation 28 th research Institution,Nanjing 210001,China)
出处
《软件导刊》
2018年第11期40-44,共5页
Software Guide
关键词
云计算
资源调度
负载均衡
负载预测
cloud computing
resource scheduling
load balancing
load prediction
