首页> 外文会议>Parallel and Distributed Computing and Systems >SANTOPIA: SHARED-DISK FILE SYSTEM FOR STORAGE CLUSTER
【24h】

SANTOPIA: SHARED-DISK FILE SYSTEM FOR STORAGE CLUSTER

机译:圣地亚哥:存储磁盘共享磁盘文件系统

获取原文

摘要

Recently, there have been large storage demands for manipulating multimedia data such as image and video. To resolve tremendous storage demands, one of the major researches is the SAN (Storage Area Network) that provides the local file requests directly from shared disk storage and also eliminates the server bottlenecks to performance and availability. The SAN also improves the network latency and bandwidth through new channel interface like FC (Fibre Channel). The FC is capable of maintaining several simultaneous gigabit data transfer. But to make use of the efficient storage network like SAN, a traditional file system is not adaptable in terms of scalability, availability and consistency issues. Therefore, we propose the new shared-disk file system, the so-called SANtopia File System, for share disk storage that manipulates the large-scale inode objects and provides a key cluster enabling technology for Linux, helping to bring the scalability, availability, and load balancing benefits of clustering to Linux. In this paper, we describe the architecture and design issues of a shared-disk file system for shared disk storage and provide the efficient bitmap, the extent-based semi flat structure and the two-phase directory structure of using Extendible Hashing. And we also present a cache coherence protocol using buffer forwarding scheme to maintain the efficient metadata consistency. At last, we evaluate the performance in terms of average response time and I/O rate.
机译:近来,对处理诸如图像和视频之类的多媒体数据有大量的存储需求。为了解决巨大的存储需求,SAN(存储区域网络)是一项主要研究,它可以直接从共享磁盘存储中提供本地文件请求,并且还消除了服务器在性能和可用性方面的瓶颈。 SAN还通过新的通道接口(例如FC(光纤通道))改善了网络延迟和带宽。 FC能够维持多个同时的千兆位数据传输。但是要利用像SAN这样的高效存储网络,传统的文件系统就可伸缩性,可用性和一致性问题而言是无法适应的。因此,我们提出了一种新的共享磁盘文件系统,即所谓的SANtopia文件系统,用于共享磁盘存储,该存储操作了大型inode对象并为Linux提供了关键的集群启用技术,从而有助于带来可扩展性,可用性,以及集群到Linux的负载平衡优势。在本文中,我们描述了用于共享磁盘存储的共享磁盘文件系统的体系结构和设计问题,并提供了使用可扩展哈希的高效位图,基于范围的半平面结构和两阶段目录结构。并且我们还提出了使用缓冲区转发方案来保持有效的元数据一致性的缓存一致性协议。最后,我们根据平均响应时间和I / O速率评估性能。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号