An Efficient One-to-One Piggybacking Design for Distributed Storage Systems

被引:20
作者
Li, Guiyang [1 ]
Lin, Xing [2 ]
Tang, Xiaohu [2 ]
机构
[1] Sichuan Normal Univ, Coll Comp Sci, Chengdu 610101, Peoples R China
[2] Southwest Jiaotong Univ, Informat Secur & Natl Comp Grid Lab, Chengdu 610031, Peoples R China
基金
中国国家自然科学基金;
关键词
Distributed storage system; piggybacking; MDS code; systematic node; parity node; REGENERATING CODES;
D O I
10.1109/TCOMM.2019.2941933
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
As a kind of erasure code, piggybacking has been applied in practice since it can significantly reduce the repair bandwidth of distributed storage systems. Currently, several efficient piggybacking designs have been proposed. In this paper, we propose a more efficient "one-to-one" piggybacking design (OOP) to further reduce the repair bandwidth. Different from the existing piggybacking designs, OOP adopts a simple encoding principle that one parity node only piggybacks symbols from one substripe. Particularly, OOP takes into account the efficient repair of systematic nodes and parity nodes simultaneously. It is shown that for OOP design, the optimal number of substripes is (root r - 1 + r - 1), the average repair bandwidth ratio of systematic nodes can be as low as 2 root r-1+1/2 root r-1+r, and the average repair bandwidth ratio of parity nodes reaches (root r-1+1/r + (r-1)(2)-root(r-1)(3)/rk). In contrast to the existing piggybacking designs, OOP can further reduce the repair bandwidth of both system nodes and parity nodes.
引用
收藏
页码:8193 / 8205
页数:13
相关论文
共 25 条
[1]  
Alrabiah O., 2019, ARXIV190105112
[2]  
[Anonymous], HDFS RAID
[3]  
[Anonymous], 2017, APACHE HADOOP 3 0 0
[4]  
Bhagwan R, 2004, USENIX ASSOCIATION PROCEEDINGS OF THE FIRST SYMPOSIUM ON NETWORKED SYSTEMS DESIGN AND IMPLEMENTATION (NSDI'04), P337
[5]  
Calder B, 2011, SOSP 11: PROCEEDINGS OF THE TWENTY-THIRD ACM SYMPOSIUM ON OPERATING SYSTEMS PRINCIPLES, P143
[6]   Network Coding for Distributed Storage Systems [J].
Dimakis, Alexandros G. ;
Godfrey, P. Brighten ;
Wu, Yunnan ;
Wainwright, Martin J. ;
Ramchandran, Kannan .
IEEE TRANSACTIONS ON INFORMATION THEORY, 2010, 56 (09) :4539-4551
[7]  
Fikes A., 2010, Storage architecture and challenges
[8]  
Ghemawat S., 2003, Operating Systems Review, V37, P29, DOI 10.1145/1165389.945450
[9]   Rapid Recovery for Systems with Scarce Faults [J].
Huang, Chung-Hao ;
Peled, Doron ;
Schewe, Sven ;
Wang, Farn .
ELECTRONIC PROCEEDINGS IN THEORETICAL COMPUTER SCIENCE, 2012, (96) :15-28
[10]  
Kubiatowicz J, 2000, ACM SIGPLAN NOTICES, V35, P190, DOI 10.1145/384264.379239