期刊文献+

一种基于授权机制的分布式文件系统小文件访问优化策略 被引量:1

Optimization Strategy Based on Delegation Mechanism for Distributed File System in Massive Small File Access Patterns
下载PDF
导出
摘要 针对分布式文件系统应用于海量小文件访问模式时,元数据请求过多导致系统性能下降的问题,提出了客户端元数据缓存授权机制的解决方案.客户端从服务器读取元数据时,申请相应类型的授权,服务器分析请求并决定是否授予此客户端所访问的元数据的授权.若客户端成功获取授权,则将其与本地缓存的元数据相关联,作为缓冲有效性的凭证.当再次访问本地缓存的元数据时,若有相关授权,则可以直接从本地获取元数据,无需向服务器发送缓存数据的有效性验证RPC.仿真实验表明,文中的方法有效的降低了客户端发送元数据请求RPC的数量,节省了宝贵的网络带宽资源,降低了元数据服务器的负载. Aiming at the system performance degradation problems of distributed file system in massive small file access patterns, designed and implemented a delegation mechanism for meta data cache in client .The client apply some kinds of delegations when getting meta data from the server. The meta data server checked and decided ecide whether to grant the delegation or not. Once the client getting the delegation which is stored in the local memory, the meta data getting operation can be unnecessary to sent RPC request. The simulation result shows that our method in this paper can reduce the number of the RPC request effectively, and aving valuable network bandwidth resources, reducing the metadata server load.
作者 曹风华
出处 《计算机系统应用》 2013年第7期183-186,176,共5页 Computer Systems & Applications
关键词 分布式文件系统 海量小文件 元数据缓存 授权机制 pNFS distributed file system massive small files metadata cache authorization mechanism pNFS
  • 相关文献

参考文献10

  • 1Hilderbrand D, Ward L, Honeyman P. Large files, small writes,and pNFS. Proc. of the 20th Annual International Conference on Supercomputing. 2006:116-124. 被引量:1
  • 2Cams P, Lang S, Ross R, Yannur MV, Kunkel J, Ludwig T. Small-file access in parallel file systems. IPDPS. 2009. 被引量:1
  • 3Cluster File Systems Inc. Lustre: A Scalable, High-Performance File System. www.lustre.org, 2002. 被引量:1
  • 4Tomkins A, Patterson RH, Gibson G Informedmulti-processprefetching and caching. Proe. of the 1997 ACM SIGMET- RICS Conference on Measurement and Modeling of Computer Systems. Seattle, Washington, 1997: 100-114. 被引量:1
  • 5Patterson RH, Gibson GA, Ginting E. Informed prefetching and caching. 15th ACM Symposium on Operating System Principles. 1995. 被引量:1
  • 6Anderson TE, Dahlin IV[D, Neefe JM, Patterson DA, Roselli DS, Wang RY. Sel'ver-less network file systems. ACM Trans. on Computer Systems, ACM, February1996,14( 1):41-79. 被引量:1
  • 7Hendricks J, Sambasivan RR, Sinnamohideen S, Ganger GR. Improving small file performance in obje.'t-bastxl storage. Technical Report 06-104, Parallel Data Laboratory, CarnegieMellon University, 2006. 被引量:1
  • 8Patil S, Gibson G Scale and Concurrency of GIGA+: File System Directories with Millions of Files. Proc. of the 9th USENIX Conference on File and Storage Technologies (FASTI 1). San Jose CA, February 2011. 被引量:1
  • 9https://btrfs.wiki.kemel.org/.April 24,2012. 被引量:1
  • 10Radkov P, Yin L, Goyal P, Sarkar P, Shenoy P. A Performance Comparison of NFS and iSCSI for IP- Networked Storage. Proc. of the 3rd USENIX Conference on File and Storage Technologies. March 31, 2004, San Francisco, CA. 被引量:1

同被引文献4

引证文献1

二级引证文献8

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部