基于COG-OS框架利用SMART预测云计算平台的硬盘故障.pdfVIP

Journal of Computer Applications 计算机应用，2014，34(1)：31—35，188 ISSN 1001—908l CODEN JYIIDU 2014．01．10 http：／／www．joca．cn 文章编号：1001—9081(2014)01-0031-05 doi：10．11772／j．issn．1001-9081．2014．01．0031 基于COG-OS框架利用SMART预测云计算平台的硬盘故障 宋云华1’r，柏文阳1’2，周琦3 (1．南京大学计算机科学与技术系，南京210023； 2．计算机软件新技术国家重点实验室(南京大学)，南京210023； 3．阿里云计算有限公司飞天一结构化数据服务，杭州311121) (}通信作者电子邮箱waterbaby66@aliyun．com) 摘要：针对云计算平台的硬盘不可靠问题，提出基于带过采样的COG(COG—OS)框架，利用硬盘自我监测分析 和报告技术(SMART)日志预测故障硬盘。首先采用DBScan或K-means聚类算法将无故障硬盘样本划分成多个不相 交子集；再与故障硬盘样本结合，采用少量样本合成过采样技术(SMOTE)使整体样本集趋于平衡；最后采用LIBSVM 分类算法预测故障硬盘。调整参数，将COG—OS与SMOTE+支持向量机(SVM)的预测性能相比较，实验结果表明该 方法具有可行性。当采用K-means方法划分无故障盘样本，并采用径向基函数(RBF)内核的LIBSVM方法预测故障 盘时，COG．OS改善了SMOTE+SVM对故障硬盘的预测查全率和整体性能。 关键词：COG—OS框架；自我监测分析和报告技术；K-均值；少量样本合成过采样技术；LIBSVM；支持向量机 中图分类号：TP306．3；TP309．1 文献标志码：A Prediction on hard disk陆lure of cloud computing framework by using SMART on COG-OS framework SONG Yunhual’”，BAI Wenyangk 2，ZHOU Qij (1．Department of Computer Science and Technology,Na咖ng University,Nanjing Jiangsu 210023，China； 2．State Key Laboratoryfor Novel So．矗ware Technology,Nanjing University,Nanjing J谊ngsu 210023，China； 3．Flying-Structured Data Services，Alibaba Cloud Computing Limited，Hangzhou Zhejiang 31 1 121，China) Abstract：The hard disk of cloud computing platform is not reliable．This paper proposed to use Self-Monitoring Analysis and Reporting Technology(SMART)log to predict hard disk failure based on Classification using lOcal clusterinG with Over- Sampling(COG-OS)framework．First，faultless hard disks were divided into multiple disjoint sample subsets by using DBScan or K-means clustering algorithm．And then these subsets and another sample set of faulty hard disks were mixed，and Synthetic Minority Over．sampling TEchnique(SMOTE)Was used to make the overall sample set tend to balance．At last， faulty hard disks Was predicted by using LIBSVM classification algorithm．The experimental results show that the method iS fe∞ible．COG．OS improves SMOTE+Support Vector Machine(SVM)on faulty hard disks’recall a