期刊文献+

基于用户态文件系统的共享访问与性能提升研究 被引量:4

Research on Sharing Access and Performance Improvement Based on User-Space File System
下载PDF
导出
摘要 研究并实现面向NVMe SSD的用户态高性能共享式文件系统(UHSFS),并提出简单弹性的数据布局技术、多粒度IO队列弹性分离技术以及多用户共享内存架构,这些关键技术能够显著地提升元数据的操作性能和IO处理能力,并且实现用户态文件系统的共享访问。实验结果表明,与UNFS,NVFUSE,BLOBFS和BLUEFS等用户态文件系统相比,UHSFS的元数据操作性能和Filebench综合负载性能最优;IO队列分离技术能够大幅度地提升吞吐量;与F2FS,XFS,Ext4等内核态文件系统相比,UHSFS同样表现出性能优势。 This paper designs UHSFS,a user-space high-performance shared file system for NVMe SSD,proposing simple elastic data layout technology,elastic separation of multiple IO queues technology,and multiuser shared memory architecture.The metadata operation performance and IO processing ability are significantly improved,and the shared access of user-space file system is realized.Experiments show that UHSFS has the best metadata operation performance and Filebench workload performance,compared with user-space file systems including UNFS,NVFUSE,BLOBFS,and BLUEFS;queue separation technology can greatly improve IOPS and throughput;UHSFS also has better performance compared with kernel-space file systems including F2FS,XFS and Ext4.
作者 杨洪章 杨雅辉 孙广宇 YANG Hongzhang;YANG Yahui;SUN Guangyu(School of Software&Microelectronics,Peking University,Beijing 102600;School of Electronics Engineering and Computer Science,Peking University,Beijing 100871)
出处 《北京大学学报(自然科学版)》 EI CAS CSCD 北大核心 2020年第6期1020-1030,共11页 Acta Scientiarum Naturalium Universitatis Pekinensis
基金 国家重点研发计划(2018YFB1003304) 国家自然科学基金(61672283) 江苏省工业和信息产业转型升级专项资金资助。
关键词 用户态 文件系统 NVMe协议 固态硬盘 user-space file system NVMe protocol solid state disk
  • 相关文献

参考文献1

二级参考文献44

  • 1Chen F, Lee R, Zhang X. Essential roles of exploit internal parallelism of flash memory based solid state dri in high-speed data processing [C] //Proc of the 17th IEEE lng ves Int Symp on High Performance Computer Architecture. Piscataway, NJ: IEEE, 2011: 266-277.
  • 2Hu Y, Jiang H, Feng D, et al. Performance impact and interplay of SSD parallelism through advanced commands, allocation strategy and data granularity [C] //Proc of the Int Conf on Supercomputing. New York: ACM, 2011:96-107.
  • 3Josephson W K, Bongo L A, Flynn D, et al. DFS: A file system for virtualized flash storage [C] //Proc of the 8th USENIX Conf on File and Storage Technologies. Berkeley: USENIX Association, 2010:85-99.
  • 4Zhang Y, Arulraj L P, Arpaci-Dusseau A C, et aI. De- indirection for flash-based SSDs with nameless writes [C] // Proe of the 10th USENIX Conf on File and Storage Technologies. Berkeley= USENIX Association, 2012: 1-16.
  • 5Nellans D, Zappe M, Axboe J, et al. ptrim() + exists() Exposing new FTL primitives to applications [C/OL]. 2011 [2012-09-01]. http://david, nellans, org/files/NVMW-2011. pdf.
  • 6Ouyang X, Nellans D, Wipfel R, et al. Beyond block I/O: Rethinking traditional storage primitives [C] //Proe of the 17th IEEE Int Symp on High Performance Computer Architecture. Piseataway, NJ: IEEE, 2011:301-311.
  • 7Prabhakaran V, Rodeheffer T L, Zhou L. Transactional flash [C] //Proe of the 8th USENIX Conf on Operating Systems Design and Implementation. Berkeley: USENIX Association, 2008:147-160.
  • 8On S, Xu J, Choi g, etal. Flag commit: Supporting efficient transaction recovery on flash-based DBMSs [J]. IEEE Trans on Knowledge and Data Engineering, 2011, 24 (9) : 1624- 1639.
  • 9Hsu P, Chang Y, Huang P, et al. A version-based strategy for reliability enhancement of flash file systems [C] //Proc of the 48th Design Automation Conf. New York: ACM, 2011: 29-34.
  • 10Ranganathan P, Chang J. (Re) Designing data centric data centers[J]. IEEEMicro, 2012, 32(1):66-70.

共引文献53

同被引文献33

引证文献4

二级引证文献4

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