Research on persistent memory file system optimization

Jianquan Zhang, D. Feng, Jingning Liu, Lei Yan, Zheng Zhang
{"title":"Research on persistent memory file system optimization","authors":"Jianquan Zhang, D. Feng, Jingning Liu, Lei Yan, Zheng Zhang","doi":"10.1109/ICCSNT.2017.8343673","DOIUrl":null,"url":null,"abstract":"Emerging byte-addressable non-volatile memory offers high performance as DRAM and data persistency like disks. In order to use the NVM in a reasonable manner, existing storage systems put the NVM on memory bus and design a specialized file system to take advantage of the characteristics of NVM. However, the current NVM-based file systems use the page table to provide direct access in NVM, which cause performance overhead in page fault. In this paper, to address this issue, we propose a unified mapping structure based memory-mapped file system. It combines the advantages of memory page table and file data index, and uses the page table structure based data index as the unified address structure. By using this method, we can efficiently implement the mapping and release of resources, also, we can effectively reduce the overhead of page fault. By using the memory simulator, we evaluate our file system in various workloads. Evaluation results show that, our unified mapping structure based memory-mapped file system effectively improves the access performance and reduces space overhead during the accessing process.","PeriodicalId":163433,"journal":{"name":"2017 6th International Conference on Computer Science and Network Technology (ICCSNT)","volume":"38 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2017-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2017 6th International Conference on Computer Science and Network Technology (ICCSNT)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICCSNT.2017.8343673","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

Abstract

Emerging byte-addressable non-volatile memory offers high performance as DRAM and data persistency like disks. In order to use the NVM in a reasonable manner, existing storage systems put the NVM on memory bus and design a specialized file system to take advantage of the characteristics of NVM. However, the current NVM-based file systems use the page table to provide direct access in NVM, which cause performance overhead in page fault. In this paper, to address this issue, we propose a unified mapping structure based memory-mapped file system. It combines the advantages of memory page table and file data index, and uses the page table structure based data index as the unified address structure. By using this method, we can efficiently implement the mapping and release of resources, also, we can effectively reduce the overhead of page fault. By using the memory simulator, we evaluate our file system in various workloads. Evaluation results show that, our unified mapping structure based memory-mapped file system effectively improves the access performance and reduces space overhead during the accessing process.
持久内存文件系统优化研究
新兴的字节可寻址非易失性存储器提供了像DRAM一样的高性能和像磁盘一样的数据持久性。为了合理地使用NVM,现有的存储系统将NVM放在内存总线上,并设计专门的文件系统来利用NVM的特性。但是,当前基于NVM的文件系统使用页表在NVM中提供直接访问,这导致了页错误带来的性能开销。为了解决这一问题,本文提出了一种基于统一映射结构的内存映射文件系统。它结合了内存页表和文件数据索引的优点,采用基于页表结构的数据索引作为统一的地址结构。利用该方法,可以有效地实现资源的映射和释放,也可以有效地减少页面故障带来的开销。通过使用内存模拟器,我们可以在各种工作负载下评估文件系统。评估结果表明,基于统一映射结构的内存映射文件系统有效地提高了访问性能,降低了访问过程中的空间开销。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
自引率
0.00%
发文量
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信