Font Size: a A A

Optimization Of Write Performance And Scalability Of Disk Arrays For Correcting Triple Disk Failures

Posted on:2018-05-09Degree:MasterType:Thesis
Country:ChinaCandidate:Y B JiangFull Text:PDF
GTID:2428330590977647Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
In the information explosion era,with the sharp increasing amount of user data and requirements of storage devices,concurrent multiple disk failures are not rare.Triple Disk Failure Tolerant arrays(3DFTs)gain much popularity due to their high reliability and low monetary cost.In large data centers,erasure code is one of the most efficient ways to protect user data with low monetary cost.One class of erasure codes is called Maximum Distance Separable(MDS)codes,which aims to offer data protection with minimal storage overhead.However,existing Triple Disk Failure Tolerant arrays(3DFTs)based on MDS codes suffer from low single write performance and insufficient scalability.Single write performance is low,because the corresponding codes have high computational cost and low encoding performance.Scalability issue is a challenging one because existing approaches have to cost plenty of migration I/Os on balancing the data distribution according to the complex layout of erasure codes.To address the write performance problem,in this paper,we propose a novel MDS coding scheme called EH-Code.EH-Code is an extension of H-Code and has three different parities,horizontal,diagonal and anti-diagonal parities,which can tolerate concurrent disk failures of any triple disks.Our mathematical analysis shows that EH-Code offers optimal storage efficiency and encoding computational complexity.Specifically,compared to STAR code,Triple-Star code and Cauchy-RS codes,EH-Code can improve the single write performance by up to 16.13%,14.53% and 26.27%,respectively.To address the scalability problem,we propose a novel Balanced Data Redistribution scheme(BDR)to accelerate the scaling process.BDR scheme migrates proper data elements according to a global point of view on a stripe set,which guarantees uniform data distribution and a small number of data movements.To demonstrate the effectiveness of BDR,we conduct several evaluations and simulations.The results show that,compared to typical RAID scaling approaches like Round-Robin(RR),Stripe-based Data Migration Scheme(SDM)and RS6,BDR scheme reduces the scaling I/Os by up to 77.45%,which speeds up the scaling process of3 DFTs by up to 4.17×,3.31×,3.88×,respectively.
Keywords/Search Tags:3DFTs, Erasure Code, Write Performance, Scalability, Performance Evaluation
PDF Full Text Request
Related items