數據中心的日漸成長, 所使用的資源與硬體設備, 也越難保持在一樣的品牌, 效能與特性上.現今的數據中心已經使用各種品牌的設備, 混合使用高階與低階的硬體, 在足服務品質之餘讓整體的花費降到最低. 然而, 使用異質性的環境也讓管理變的更加複雜, 尤其是在各種資源的管理與分配. 本篇文章, 我們著重在塊狀資源資源的分配, 並且考量同時使用HDD與SSD的環境.我們針對不同的儲存裝置進行實驗, 以量化每個裝置的在不同的存取模式的表現. 並且將資源分配問題轉化為整數線性規畫 (ILP)問題. 最後, 我們藉由實作一個異質性的儲存系統, 其中包含HDD, SSD與 iSCSI硬碟來評估我們的解決方案, 並且可以減少35%~81%的存取響應時間.

As the scale of datacenter continues to grow, it is hard to keep servers homogenous, with the same hardware and performance characteristics. Today’s datacenters commonly operates on several generations of servers from multiple vendors, and mix both high-end and low-end devices together to deliver service quality requirement with lowest cost. However, the heterogeneous environment also complicates the management of the datacenters, especially in terms of resource allocation. In this paper, we focus on the resource allocation of a tightly unified block-level storage with SSD and HDD. We conduct experiments to quantify the performance of difference access patterns on each type storage devices. Then formulate our resource allocation problem into a ILP (Integer Linear Programming), and proposed data migration algorithms based on the observations.We evaluate our solution by implementing a heterogenous storage consist of HDD, SDD and iSCAII HDD, and show the data access response
time can be reduced by 35%  81%.

[1] H. Kopka and P. W. Daly, A Guide to LATEX, 3rd ed. Harlow, England:
Addison-Wesley, 1999.
[2] Xiaonan Zhao, Zhanhuai Li and Leijie Zeng. “A Hierarchical Storage
Strategy Based on Block-Level Data Valuation”. Fourth International
Conference on Networked Computing and Advanced Information Management,
IEEE, 2008, Page(s): 36–41
[3] Jin Ren and Qing Yang. “I-CASH: Intelligently Coupled Array of SSD
and HDD”. High Performance Computer Architecture (HPCA), IEEE,
2011, Page(s): 278–289
[4] Dushyanth Narayanan, Eno Thereska, Austin Donnelly, Sameh Elnikety
and Antony Rowstron. “Migrating Server Storage to SSDs: Analysis of
Tradeoffs”. European conference on Computer systems (EuroSys), ACM,
2009, Page(s): 145–158
[5] Gokul Soundararajan, Vijayan Prabhakaran, Mahesh Balakrishnan and
Ted Wobber. “Extending SSD Lifetimes with Disk-Based Write Caches”.
USENIX conference on File and storage technologies (FAST), ACM, 2010,
Page(s): 8–8
[6] Sang-Hoon Kim, Dawoon Jung, Jin-Soo Kim and Seungryol Maeng.
“HeteroDrive: Reshaping the Storage Access Pattern of OLTP Workload
Using SSD”. International Workshop on Software Support for Portable
Storage (IWSSPS), 2009, Page(s):13–17.
[7] bitprog, Solve binary integer programming problems.
http://www.mathworks.com/help/toolbox/optim/ug/bintprog.html.
MATLAB.
[8] Chris Ruemmler and JohnWilkes. An introduction to disk drive modeling.
IEEE Computer Society Press Los Alamitos, 1994, Page(s): 17–18
[9] UMass Trace Repository. http://traces.cs.umass.edu/index.php/Storage/Storage.
[10] KHACHIAN, L G. Polynomial algorithms in linear programming.
Zhurnal Vychislitel’noi Matematiki i Matematicheskoi Fiziki, 1980, Vol.
20, Page(s) 51-68
[11] hdparm - get/set SATA/IDE device parameters.
http://linux.die.net/man/8/hdparm.
[12] Ying Chen, information valuation for information lifecycle management,
Proceedings of international Conference on Autonomic Computing (ICAC
05), June 2005, pp. 135-146.
[13] Muzhou Xiong, Hai Jin, Song Wu, Information Lifecycle Management
in Multi-Gird Environment, Proceeding of the fourteenth national conference
on information and storage techniques, September 2006, pp. 175-
184.
[14] ILM Definition and Scope an ILM framework, July,2004.
http://www.snia.org/forums/dmf/programs/ilmi/DMF-ILMVision2.4.pdf.
[15] J. A. Nelder and R. Mead. “A simplex method for function minimization”.
The Computer Journal, 1965, Page(s): 308-313.