Please use this identifier to cite or link to this item: http://hdl.handle.net/11455/19971
標題: 具備負載平衡之混合式儲存裝置的資料快取方法
A Load-Balancing Data Caching Scheme for Hybrid Storage
作者: 羅志誠
Luo, Jhih-Cheng
關鍵字: 混合式儲存裝置
hybrid storage
資料快取機制
固態硬碟
傳統硬碟
data caching policy
solid state disks
hard disks
出版社: 資訊科學與工程學系所
引用: [1] P. Cao and S. Irani, "Cost-Aware WWW Proxy Caching Algorithms," in Proc. USENIX Symp. Internet Technologies and Systems, 1997. [2] X.-Y. Hu, E. Eleftheriou, R. Haas, I. Iliadis and R. Pletka, "Write Amplification Analysis in Flash-Based Solid State," in Proceedings of SYSTOR, 2009. [3] T. Kgil and T. Mudge, "FlashCache: A NAND Flash Memory File Cache for Low Power Web Servers," in Proceeding CASES ''06 Proceedings of the 2006 international conference on Compilers, architecture and synthesis for embedded systems, 2006. [4] Y. Kim and J. Kim, "DAC: A Device-Aware Cache Management Algorithm for Heterogeneous Mobile Storage Systems," IEICE TRANSACTIONS on Information and Systems, VOL.E91-D, NO.12, 2008. [5] Y. Kim and J. Kim, "Device-aware cache replacement algorithm for heterogeneous mobile storage devices," in ICESS ''07 Proceedings of the 3rd international conference on Embedded Software and Systems, 2007. [6] Y. Kim, A. Gupta, B. Urgaonkar, P. Berman and A. Sivasubramaniam, "HybridStore: A Cost-Efficient, High-Performance Storage System Combining SSDs and HDDs," in IEEE 19th Annual International Symposium on Modelling, Analysis, and Simulation of Computer and Telecommunication Systems, 2011. [7] Y.-J. Kim and J.-H. Kim, "ARC-H: Adaptive replacement cache management for heterogeneous storage devices," Journal of Systems Architecture Volume 58, Issue 2, 2012. [8] S.-Y. Liao, “A Data Caching Scheme for Hybrid Hard Drives,” 2010. [9] Y. Liu, J. Huang, C. Xie and Q. Cao, "RAF: A Random Access First Cache Management to Improve SSD-based Disk Cache," in Fifth IEEE Internatinoal Conference on Networking, Architecture, and Storage, 2010. [10] Y. Oh, J. Choi, D. Lee and S. H. Noh, "Caching less for better performance: Balancing cache size and update cost of flash memory cache in hybrid storage systems," in 10th USENIX Conference on File and Storage Technologies, 2012. [11] "DiskSim," http://www.pdl.cmu.edu/DiskSim/. [12] "Intel Smart Response Technology," http://www.intel.com.tw/content/www/tw/zh/architecture-and-technology/smart-response-technology.html. [13] "MAXTOR ATLAS10K4 146G, 10000rpm SCSI," http://ovahldy.blogspot.com/2011/03/recent-diskmodels-for-disksim.htm. [14] "PCMark7," http://www.futuremark.com/benchmarks/pcmark.
摘要: 混合式硬碟是由傳統硬碟與快閃記憶體(NAND Flash memory)所組成,其主要想法是利用快閃記憶體的高存取效能和低耗電以及傳統硬碟的低價位和大容量,企圖在效能與價格之間取得平衡,利用較小容量的快閃記憶體做為傳統硬碟之快取以提昇I/O效能,同時在成本考量下亦能擁有較大的儲存空間。 先前有研究提出應用於混合式硬碟之資料快取機制“Prize Caching”,藉由分析資料的存取特性來決定哪些資料應該擺放在快閃記憶體中做為快取資料。本篇論文改善Prize Caching的方法,提出LBPC (Load-Balanced Prize Caching),考量不同儲存裝置之負載量,首先會依據門檻值判斷資料的循序性,將長度較長也就是較具循序性的資料優先過濾掉並交由傳統硬碟服務,而隨機存取的資料則交由快閃記憶體來負責,並且在每隔一段時間後將週期性地估算出新的門檻值,使得傳統硬碟與快閃記憶體的負載量能盡量達到平衡。 本論文使用基於DiskSim所開發之模擬混合式硬碟的模擬器,探討在實作上所遇到的相關議題並改善LBPC機制以提昇I/O存取效能。根據模擬之實驗結果可以證明LBPC快取策略能夠降低request的平均反應時間,提昇系統的整體效能。
Hybrid disk drives are composed of traditional hard disk and NAND Flash memory. The main idea of hybrid storage is to exploit the advantages of both devices, such as the high access performance and the low power consumption of NAND Flash, and the low price and huge storage capacity of hard disks. It attempts to trade-off between the cost and the performance, so it usually uses small size of NAND flash memory as the cache of hard disks to improve I/O performance. Therefore, it has not only high I/O performance but also a large storage space with a reasonable cost. Previously, researchers had proposed a data caching scheme for hybrid disk drives called “Prize Caching”. It decides whether data should be cached into NAND Flash by analyzing the feature and the access history of data. In this thesis, we enhance Prize Caching and propose LBPC (Load-Balanced Prize Caching). LBPC will take the load of different storage devices into consideration. At first, it detects sequential reference by comparing with the random threshold. The data whose length is too long will be considered as sequential reference and served by hard disks. On the other hands, the random reference will be served by NAND flash. In our design, in order to make the load balance between hard disks and NAND flash, the random threshold will be adjusted periodically. We implement LBPC policy onto a hybrid drive simulator which is based on DiskSim. Experimental results show that our scheme can reduce the average response time of requests and improve the overall performance.
URI: http://hdl.handle.net/11455/19971
其他識別: U0005-1408201310140200
文章連結: http://www.airitilibrary.com/Publication/alDetailedMesh1?DocID=U0005-1408201310140200
Appears in Collections:資訊科學與工程學系所

文件中的檔案:

取得全文請前往華藝線上圖書館



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.