跳到主要內容

臺灣博碩士論文加值系統

(3.236.110.106) 您好!臺灣時間:2021/07/29 16:17
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果 :::

詳目顯示

: 
twitterline
研究生:劉美蘭
研究生(外文):Liou,Meilan
論文名稱:基於混合型儲存設備下針對雲端多媒體應用之動態資料遷移
論文名稱(外文):Dynamic Data Migration Of Multimedia Applications On HDD/SSD Based Clouds
指導教授:張榮貴張榮貴引用關係
指導教授(外文):Chang, Rong-Guey
口試委員:陳鵬升黃元欣薛智文
口試委員(外文):Chen, Peng-ShengHwang, Yuan-ShinHsueh, Chih-Wen
口試日期:2012-07-23
學位類別:碩士
校院名稱:國立中正大學
系所名稱:資訊工程研究所
學門:工程學門
學類:電資工程學類
論文種類:學術論文
論文出版年:2012
畢業學年度:100
語文別:英文
論文頁數:34
中文關鍵詞:多媒體混合型儲存設備
外文關鍵詞:MultimediaData migration
相關次數:
  • 被引用被引用:0
  • 點閱點閱:304
  • 評分評分:
  • 下載下載:9
  • 收藏至我的研究室書目清單書目收藏:1
為了提高多媒體應用在雲端上的性能以及節省其功耗,我們設計並實作一個基於傳統硬碟與固態硬碟的架構,來解決這些議題。
其中一項就是如何解決大量多媒體數據在雲端網路所造成的壅塞。然而,這個問題跟多媒體的儲存設備是
非常密切相關的。所以我們基於傳統硬碟和固態硬碟的特點來提供優越的頻寬,進而解決這個問題。

在利用資料遷移來增進混合儲存裝置吞吐量的研究中,均沒有掌握到熱門檔案變化的曲線,只著重在發揮固態硬碟的優勢,
決策如何搬移檔案。在系統運作時,對該在什麼時候執行資料搬遷完全沒有細節的描述與規畫。

在本篇論文中,我們呈現了一個在傳統硬碟和固態硬碟之間動態資料遷移的方法,稱為IDDM。
IDDM提供了一個虛擬化的平台,並提出藉由排程虛擬機器找到排程空檔來執行資料搬遷,使得性能提高且不會錯過即時系統的截止期。
與以前的研究作比較,我們的動態資料遷移更加的敏感以及接近使用者行為的曲線,能夠提高在固態硬碟中的命中率,改善
混合儲存裝置的效能。

IDDM能夠利用系統的空檔在混合儲存裝置之間動態的搬遷檔案,且動態的變動熱門檔案的門檻值,提升了系統吞吐量高達30%。
To improve performance and save power consumption of multimedia clouds, we have designed and implemented a infrastructure to solve these issues based on HDD/SDD (Hard Disk Drive and Solid State Drive). One of our infrastructure is how to solve the congestion caused by a very large amount of multimedia data. However, this issue is very closely related to the storage device of multimedia applications. Thus to overcome this issue, we must provide bandwidth based on the characteristics of HDD and SDD.

The studies that rely on the data migration to enhance hybrid storage throughput don't grasp the popularity curves. They only focus on the advantages of SSDs to play up the path of the decision to move files, but not to mention when the implementation of the data migration in the system cooperation.

In this paper, we present a dynamic migration mthod called IDDM (Idle-driven Data Migration) to move data between HDD and SDD.
IDDM provides a virtualization platform and proposes a virtual machine scheduler to migrate data between HDD and SDD for good performance gaps without missing the real-time deadline.Compared with the previous work, our dynamic migration is more sensitive and close to the curve of the user behavior change, and enhances the SSD hit ratio to improve performance of the hybrid storage.

IDDM takes advantage of the gap on the scheduler to move files dynamically in hybrid storage and change a threshold, and finally improves system throughput by up to 30%.
1 Introduction 1
2 Related Work 4
2.1 Caching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2 Hybrid Storage . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.3 File Assignment Algorithm . . . . . . . . . . . . . . . . . . . . 6
3 Motivation 8
3.1 Rate of Change in User Interest . . . . . . . . . . . . . . . . . 8
3.2 Understand Server Utilization . . . . . . . . . . . . . . . . . . 8
3.3 CPU Scheduler . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4 IDDM 11
4.1 Architecture Overview . . . . . . . . . . . . . . . . . . . . . . 11
4.2 Design Procedure . . . . . . . . . . . . . . . . . . . . . . . . . 12
4.3 Trace Measurement . . . . . . . . . . . . . . . . . . . . . . . . 14
4.4 SEDF-GAP: Idle Time Driven . . . . . . . . . . . . . . . . . . 15
4.5 User Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
5 Experiment Evaluation 18
5.1 Simulation Model . . . . . . . . . . . . . . . . . . . . . . . . . 18
5.2 System Parameters . . . . . . . . . . . . . . . . . . . . . . . . 19
5.3 Workload Parameters . . . . . . . . . . . . . . . . . . . . . . . 20
5.4 Evaluation Result . . . . . . . . . . . . . . . . . . . . . . . . . 21
6 Conclusion 25

[1] Cisco visual networking index: Forecast and methodology. 2011-2016.
[2] Lee D. Pdc-nh: popular data concentration on nand ash and hard disk
drive. pages 196200, 2009.
[3] R. Gardner D. Gupta, L. Cherkasova and A. Vahdat. Enforcing perfor-
mance isolation across virtual machines in xen. pages 342362, 2006.
[4] B. T. Gold D. Meisner and T. F.Wenisch. Powernap: Eliminating server
idle power. 2009.
[5] A. Donnelly S. Elnikety and A. Rowstron D. Narayanan, E. Thereska.
Migrating enterprise storage to ssds: analysis of tradeos. pages 145
158, 2009.
[6] Austin Donnelly Sameh Elnikety Antony Rowstron
Dushyanth Narayanan, Eno Thereska. Migrating server storage
to ssds: analysis of tradeos. April 01-03 2009.
[7] Sedigh-Ali S Ghafoor A Liu P Hsu L Fahmi H, Latif M. Proxy servers
for scalable interactive video support. pages 5460, 2001.
[8] Zhong Hu and Jun Zhou Fengbin Li, Jun Li. Common caching replace-
ment algorithm for video-on-demand system. 2009.
[9] B. Y. Zhao H. Yu, D. Zheng and W. Zheng. Understanding user behavior
in large-scale video-on-demand systems.
[10] T. Wauters J. Famaey and F. D. Turck. On the merits of popularity
prediction in multimedia content caching. 2011.
[11] K. W. Hur J. G. Lee and Y. W. Ko. Minimizing scheduling delay for
multimedia in xen hypervisor. pages 96108, 2011.
[12] Z. Xiao L. Guo, S. Chen and X. Zhang. Disc:dynamic interleaved seg-
ment caching for interactive streaming. June 2005.
[13] P. Scheauermann L. Lee and R. Vingralek. Assignment in parallel i/o
systems with minimal variance of service time. pages 127140, 2000.
[14] Jianhui Yue Zhao Cai Lin Lin, Yifeng Zhu and Bruce Segee. Hot random
o-loading: A hybrid storage system with dynamic data migration. 2011.
[15] H. Kim M. Ryu and U. Ramachandran. Impact of ash memory on
video-on-demand storage: Analysis of tradeos. 2011.
[16] M. E. J. Newman. Power laws, pareto distributions and zipf's law. 2005.
[17] K. Fraser S. Hand T. Harris A. Ho R. Neugebauer I. Pratt P. Barham,
B. Dragovic and A. Wareld. Xen and the art of virtualization. pages
164177, 2003.
[18] K. Shin X. Zhu M. Uysal Z. Wang S. Singhal P. Padala, K.-Y. Hou and
A. Merchant. Automated control of multiple virtualized resources. 2009.
[19] J.-S. Kim S.-H. Kim, D. Jung and S. Maeng. Heterodrive: Re-shaping
the storage access pattern of oltp workload using ssd. pages 1317,
October 2009.
[20] Susie Wee Song-qing Chen, Bo Shen. Segment-based streaming media
proxy modeling and optimization. pages 243256, 2006.
[21] Don Towsley Subhabrata Sen, Jennifer Rexford. Proxy prex caching
for multimedia streams. pages 13101319, 1999.
[22] Bo Shen Sung-Ju Lee, Wei-Ying Ma. An interactive video delivery and
caching system using video summarization. pages 424435, 2002.
[23] T. Xie and D. Madathil. Sail: Self-adaptive le reallocation on hybrid
disk arrays. pages 529540, 2008.
[24] Tao Xie. Sor: A static le assignment strategy immune to workload
characteristic assumptions in parallel i/o systems. 2007.
[25] B. Li X. Qiu Y. Wu, C. Wu and F. Lau. Cloudmedia: When cloud on
demand meets video on demand. 2011.
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top