內建服務品質考量之資料回復策略的節能糾刪碼框架研究設計

A QoS-aware Data Reconstruction Strategy within an Energy-efficient Erasure Codes Framework

張乃元(Nai-Yuan Jhang)

指導教授 ： 石維寬

國立清華大學/電機資訊學院/資訊工程學系/碩士(2015年)

摘要

近年來，市面上有許多雲端儲存系統，像是：GOOGLE Drive、Dropbox、iCloud跟Sky Drive,在儲存用戶的私用資料跟系統資料的方面必須保證端儲存系統必須有可靠性跟安全性，因此雲端儲存系統為了提高資料的可靠性，市面的儲存系統大多數會使用糾錯碼機制來當作其系統的資料容錯機制，雖然這些容錯機制增加了資料的可靠性，但是也增加系統儲存成本跟耗電量，然而，目前節能資料儲存系統都是開發在RAID-base的架構，並沒有開發在使用糾錯碼的系統，根據上述，我們想要提供一個能節能並有著基於糾錯碼資料回復上面有著服務品質考量的叫做EECache-scheme儲存系統，這個方案可以減少在資料回覆時候的耗電量跟減少回應使用者需求時間。

關鍵字

糾刪碼 ； 節能 ； 資料回復

並列摘要

Recently, many applications and users rely on cloud storage services, such as
Google drive, Dropbox, iCloud and Sky drive, to store private files and system data,
and cloud storage services must thus be reliable and secure. To increase reliability,
previous studies have proposed a variety of erasure coding algorithms for data fault
tolerance for use in storage systems. Although these data fault tolerance mechanisms
increase data reliability, implementation also increases storage system costs and energy
consumption due to data redundancy. However, to date energy-efficient schemes have
only been developed based on a RAID architecture, and none have been implemented
using an erasure coding algorithm. To address this issue, this study proposes an energyaware
I/O framework with a quality-of service (QoS) aware data reconstruction
scheduler for erasure coding algorithms, called the EECache-scheme. This approach
reduces storage system energy consumption and decreases response times for user
requests when the system restores failed disks. A series of experiments show that the
proposed scheme can significantly reduce power consumption in storage systems and
reduces response times by xx% over a solution without QoS control.

參考文獻

[1] D. Ford, F. Labelle, F. I. Popovici, M. Stokely, V.-A. Truong, L. Barroso, C. Grimes, and S. Quinlan, Availability in Globally Distributed Storage Systems, in the 9th USENIX Symposium on Operating Systems Design and Implementation (OSDI ’10), 2010.

[4] H.-Y. Lin and W.-G. Tzeng, A Secure Decentralized Erasure Code for Distributed Networked Storage, in IEEE Transactions on Parallel and Distributed Systems, vol. 21, no. 11, pp. 1586-1594, 2010.

[5] H.-Y. Lin and W.-G. Tzeng, A Secure Erasure Code-Based Cloud Storage System with Secure Data Forwarding, in IEEE Transactions on Parallel and Distributed Systems, vol. 23, no. 6, pp. 995-1003, 2012.

[6] X. Luo and J. Shu, Load-Balanced Recovery Schemes for Single-Disk Failure in Storage Systems with Any Erasure Code, in the 42nd International Conference on Parallel Processing (ICPP13), pp. 552-561, 2013.

[9] T. Xie, SEA: A Striping-Based Energy-Aware Strategy for Data Placement in RAID-Structured Storage Systems, in IEEE Transactions on Computers, vol. 57, no. 6, pp. 748-761, 2008.