通信学报2011，Vol.32Issue(7)：47-56,10.
基于数据中心负载分析的自适应延迟调度算法
Adaptive delay scheduling algorithm based on data center load analysis
摘要
Abstract
Current delay scheduling algorithms, based on static waiting time threshold, can not adapt to dynamic load of a data center. To address this issue, an adaptive waiting time threshold model was proposed. Based on this model, an adap-tive delay scheduling algorithm (ADS) was designed. ADS adjusts jobs' waiting time threshold dynamically to reduce the job response time, according to the information of idle servers' arrival intensity, available network bandwidth and job running status. A prototype system using ADS was developed, on which the validation of the adaptive waiting time threshold model and the evaluation of ADS's performance were carried out It is shown that ADS outperforms previous delay scheduling algorithms in term of the job response time.关键词
云计算/延迟调度算法/数据本地性:自适应/数据中心/HadoopKey words
cloud computing/ delay scheduling algorithm/ data-locality/ adaptive/ data center/ Hadoop分类
信息技术与安全科学引用本文复制引用
金嘉晖,罗军舟,宋爱波,东方..基于数据中心负载分析的自适应延迟调度算法[J].通信学报,2011,32(7):47-56,10.基金项目
国家自然科学基金资助项目(61070161,61070158,61003257,60773103,90912002) （61070161,61070158,61003257,60773103,90912002）
国家重点基础研究发展(“973”计划)基金资助项目(2010CB328104) （“973”计划）
国家科技支撑计划课题基金资助项目(2010BAI88B03) （2010BAI88B03）
教育部博士点基金课题基金项目(200802860031) （200802860031）
江苏省自然科学基金资助项目(BK2008030) （BK2008030）
国家科技重大专项课题基金资助项目(2009ZX03004-004-04) （2009ZX03004-004-04）
江苏省“网络与信息安全”重点实验室项目(BM2003201) （BM2003201）
“计算机网络与信息集成”教育部重点实验室项目(93K-9) （93K-9）
