摘要
随着体育赛事网络直播及虚拟现实等新兴业务的出现,热门体育赛事网络直播越来越多,以视频为主的网络流量急剧增长。针对爆炸式增长的视频请求流量,提出一种适用于各种工作条件(包括I/O密集型和内存密集型)的服务器集群负载均衡算法。将视频请求业务周期定义为从辅助系统加载视频文件到将视频流量连续地传输给用户。再根据单个服务器的内存、CPU和I/O需求来平衡对于整个服务器集群的请求,从而最小化每个作业的响应时间和完成时间。将该算法与两种典型策略进行比较,结果表明该算法具有更好的性能。
With the emergence of new businesses such as live broadcast of sports events and virtual reality,there are more and more live broadcast of popular sports events,which leads to the rapid growth of video-based network traffic.This paper focuses on the explosive growth of video request traffic and proposes a server cluster load balancing algorithm for a lot of network load conditions including I/O and memory intensive loads.It defined the business cycle of video request as loading video files from the auxiliary system to continuously transmitting video traffic to the user.Then,according to the memory,CPU and I/O requirements of a single server,it balanced the requests for the whole server cluster,so as to minimize the response time and completion time of each job.Compared with another two classical load balance strategies,the results show that our algorithm has better performance.
作者
葛晓燕
Ge Xiaoyan(School of Physicl,Shangqiu College,Shangqiu 476000,Henan,China)
出处
《计算机应用与软件》
北大核心
2020年第6期166-170,共5页
Computer Applications and Software
基金
河南省高等学校重点科研项目(18A890005)。
关键词
体育赛事
体育直播
负载均衡
服务器集群
Sports event
Sport live streaming
Load balancing
Server cluster
作者简介
葛晓燕,讲师,主研领域:体育科技应用,计算机网络。
