Delta Raid To Enhance Recovery And Small-Write Performance Of Raid Storages

2016 Intl IEEE Conferences on Ubiquitous Intelligence & Computing, Advanced and Trusted Computing, Scalable Computing and Communications, Cloud and Big Data Computing, Internet of People, and Smart World Congress (UIC/ATC/ScalCom/CBDCom/IoP/SmartWorld)(2016)

引用 2|浏览2
暂无评分
摘要
This paper proposes a novel Delta redundant array of independent disks (RAID) technique that enhances small-write performance and resists disk failures of up to 75% in the RAID architecture. In order to generate parity data, three partial strips within a horizontal strip are used. To generate one parity element, Delta RAID requires access from two data disks and one parity disk. Delta RAID provides a fast recovery algorithm to correct various disk failures. Delta RAID also employs additional buffer caching replacement (BCR), which reduces the number of read and write operations by both merging small chunks into a larger chunk with buffer size and splitting a big chunk into many smaller chunks with buffer size. Compared to traditional GRID RAID(SPC, STAR), GRID RAID(SPC, EO), and GRID RAID(STAR, STAR), Delta RAID improves average small-write performance by 233%, 224%, and 176%, respectively, improves read performance by 289%, 268%, and 176%, respectively.
更多
查看译文
关键词
RAID Architecture,Erasure Codes,Small-Write
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要