{"title":"Web proxy analysis in distributed iSCSI RAIDs","authors":"Dan Zhou, Xubin He","doi":"10.1109/SSST.2004.1295653","DOIUrl":null,"url":null,"abstract":"With the popularity of Web services and e-commerce, we have a high demand on data storage. Because of the development of the Internet infrastructure and the low cost of deployment, implementing storage over IP has become a trend. While networked storage brings economy, it also raises performance concern. In a typical distributed Web system, I/O requests from clients users may be satisfied by many different parties along the I/O path from front-end clients to back-end storage systems via proxy and Web servers. To improve the performance, we need to identify the I/O distribution pattern along the path. This paper analyzes the I/O distribution along the I/O path of a distributed iSCSI RAID system by using kernel counters and profiles how the I/O requests are satisfied at different levels of the proxy server and storage system, which is useful to identify the performance bottleneck.","PeriodicalId":309617,"journal":{"name":"Thirty-Sixth Southeastern Symposium on System Theory, 2004. Proceedings of the","volume":"25 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2004-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"3","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Thirty-Sixth Southeastern Symposium on System Theory, 2004. Proceedings of the","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/SSST.2004.1295653","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 3
Abstract
With the popularity of Web services and e-commerce, we have a high demand on data storage. Because of the development of the Internet infrastructure and the low cost of deployment, implementing storage over IP has become a trend. While networked storage brings economy, it also raises performance concern. In a typical distributed Web system, I/O requests from clients users may be satisfied by many different parties along the I/O path from front-end clients to back-end storage systems via proxy and Web servers. To improve the performance, we need to identify the I/O distribution pattern along the path. This paper analyzes the I/O distribution along the I/O path of a distributed iSCSI RAID system by using kernel counters and profiles how the I/O requests are satisfied at different levels of the proxy server and storage system, which is useful to identify the performance bottleneck.