Optimizing File Transfer in Distributed Storage System Based on Reconfiguration of Storage Structure

Author:

Li Kai,Chen Shouming,Chen Yingda,Zhong Susheng,Huang Kaitong,He Minnuo,Yang Lixin,Li Bo,Yan Yi

Abstract

Abstract Distributed storage is the storage part of distributed computing. It is an easy-to-expand and virtualized storage resource pool. In order to meet the needs of large-scale storage structure applications, reliable mass data storage services are provided. Redundant data storage is needed to deal with the data unreliability caused by node failure. The lower the cost of file transmission, the less traffic generated in the network. According to the time logic, the regeneration process is divided into stages: data distribution, data encoding and decoding, data transmission and data reconstruction. The feedback mechanism can reduce the overhead of acquiring the load information of the data server node while ensuring the real-time performance. The determination of the activity factor function of the data block access frequency statistics needs to be established in the performance adaptability of the data block activity factor and the specific access times. The single-point dependency of the centralized storage system on the primary node is changed, which reduces the waiting delay of concurrent users and the average memory usage of the metadata server.

Publisher

IOP Publishing

Subject

General Physics and Astronomy

Reference14 articles.

1. Towards a framework for large-scale multimedia data storage and processing on Hadoop platform[J];Lai;Journal of Supercomputing,2014

2. Mitigation of NUMA and synchronization effects in high-speed network storage over raw Ethernet[J];González-Férez;The Journal of Supercomputing,2016

3. Region-based fault-tolerant distributed file storage system design in networks[J];Networks,2015

4. Load transfer capability analysis considering interconnection of distributed generation and energy storage system[J];Liao;International Transactions on Electrical Energy Systems,2014

5. Detecting Suspicious File Migration or Replication in the Cloud[J];Bowers;IEEE Transactions on Dependable and Secure Computing,2018

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3