基於共生系統模型所建立之儲存系統的成本及可靠度分析

簡易檢索 / 詳目顯示

回結果列表

研究生：	曾淑梅 Tseng, Shu-Mei
論文名稱：	基於共生系統模型所建立之儲存系統的成本及可靠度分析 Cost and Reliability Evaluation of Repairable Storage System by Symbiotic System Model
指導教授：	吳誠文 Wu, Cheng-Wen
口試委員:	吳文慶劉靖家李進福
學位類別：	碩士 Master
系所名稱：	電機資訊學院 - 電機工程學系 Department of Electrical Engineering
論文出版年：	2017
畢業學年度：	105
語文別：	英文
論文頁數：	40
中文關鍵詞：	內建自我測試、內建自我修復、共生計算、共生系統、在線測試、彈性系統
外文關鍵詞：	built-in self-test, built-in self-repair, symbiotic computing, symbotic system, on-line testing, resilient system
相關次數：	點閱：1 下載：0
分享至:	分享至facebook 分享至twitter

查詢本校圖書館目錄查詢臺灣博碩士論文知識加值系統勘誤回報

伴隨著物聯網(IOT)世代的來臨，全球所產生的電子資料量呈現指數型成長速度，使得我們對於記憶體以及儲存裝置的需求也在逐步增加。因此，記憶體及儲存裝置的成本及可靠度相關議題也變得比以往來的更重要。為了能夠有效地評估儲存裝置在物聯網系統中的成本及可靠度，我們提出了一個高階共生系統模型(symbiotic system model)，以及運用在這個模型中的同儕修復機制(peer-repair mechanism)。實驗結果顯示同儕修復機制可以延長整體儲存系統大約60%的壽命(lifetime)，除此之外，我們更進一步發現同儕修復機制較適合用在一個儲存裝置彼此之間存在較大故障率(failure rate)變異(variation)的系統，像是被應用在未來的物聯網系統中，每一個節點可能分散在一個很大的面積範圍中，彼此之間錯誤的發生率也會有比較大的差異。在探索利用同儕修復機制降低成本的可能性時，我們發現不能只替換掉壞掉的節點，還需替換掉備用資源(spare resource)剩下不多的節點，實驗結果顯示在替換壞掉的節點(node)以及只剩下原來40%備用資源(spare resource)的節點時，整體的總成本可以達到最低，因為替換成本(replacement cost)是非常高的。除此之外，總的來說，同儕修復機制能夠以較低的成本達到整體系統所需的壽命要求，實驗結果也顯示與其花非常多的錢在出貨前的產品測試上，我們可以利用同儕修復機制，以較低的成本達到所要求的系統壽命。

With the advent of the internet of things (IOT), the amount of global electronic data is growing at an exponential rate, raising the demand for memories and storages. Therefore, the cost and reliability issues of storages and memories become more important than ever. To efficiently evaluate the cost and reliability of the storage nodes of an IOT system, we propose a high-level symbiotic system (SS) model, with the novel peer-repair mechanism. Experimental results show that the peer-repair mechanism can improve around 60% of the storage system’s lifetime. We also find that peer-repair is more suitable for systems with a wide variation of failure rates, which can be applied to, e.g., future IOT systems with nodes being distributed over a wide area. In exploring the possibility of cost reduction by peer-repair mechanism, experimental results show that by replacing failed nodes as well as nodes with 40% fewer spare resources, the lowest total cost can be achieved, as the replacement cost is high. In general, peer-repair can meet the same lifetime requirement with much lower cost, as compared to testing the devices and systems much more thoroughly before shipping.

摘要    i
Abstract    ii
致謝    iii
Table of Contents    iv
List of Figures    vi
List of Tables    viii
Chapter 1    Introduction    1
1.1    Increasing Demand for Storages (Memories)    1
1.2    Issues of the Internet of Things    2
1.3    Previous Work and Objective    3
1.4    Organization    4
Chapter 2    Symbiotic System Model    5
2.1    Symbiotic Machine    5
2.2    Symbiotic System with Subsystems    7
2.3    Symbiotic Storage System with Subsystems    8
Chapter 3    Implementation    11
3.1    Simulator Overview    11
3.2    Primary Machine (MP)    14
3.3    Secondary Machine (MS)    15
3.4    Stimulator Module    19
3.5    Simulation Speed    21
3.6    Simulation Setup    23
Chapter 4    Experimental Result    25
4.1    Relative Lifetime Increment with Identical Failure Rate    25
4.2    Relative Lifetime Increment with Plural Failure Rates    28
4.3    Cost Reduction with Different Replacement Policies    31
4.4    Test Cost Reduction by Peer-Repair    35
Chapter 5    Conclusion and Future Work    37
5.1    Conclusion    37
5.2    Future Work    37
References    39


                                

[1] L. Rizzatti, “Digital Data Storage is Undergoing Mind Boggling Growth,” EE Times http://www.eetimes.com/author.asp?section_id=36&doc_id=1330462.
[2] IDC, “The Digital Universe of Opportunities: Rich Data and the Increasing Value of the Internet of Things,” Apr. 2014.
[3] Micron Technology, Micron and Industry Analyst, 2016 http://files.shareholder.com/downloads/ABEA-45YXOQ/0x0x875021/4BEAA02E-BBC2-402C-A51D-B3B2C6B8C3D4/Winter_Analyst_Day_2016.pdf.
[4] A. Shehabi, S. Smith, D. Sartor, R. Brown, M. Herrlin, J. Koomey, E. Masanet, N. Horner, I. Azevedo, W. Lintner, “United States Data Center Energy Usage Report,” Jun. 2016 https://eta.lbl.gov/sites/all/files/publications/lbnl-1005775_v2.pdf.
[5] Statista, "Global smartphone shipments forecast from 2010 to 2020," https://www.statista.com/statistics/263441/global-smartphone-shipments-forecast/.
[6] Statista, "Forecast unit shipments of tablets worldwide from 2010 to 2020," https://www.statista.com/statistics/269912/worldwide-tablet-shipments-forecast/.
[7] Ericsson, “CEO to shareholders: 50 billion connections 2020,” https://www.ericsson.com/thecompany/press/releases/2010/04/1403231.
[8] Cisco, "The internet of things: how the next evolution of the internet is changing everything," https://www.cisco.com/c/dam/en_us/about/ac79/docs/innov/IoT_IBSG_0411FINAL.pdf
[9] C.-W. Wu, B.-Y. Lin, H.-W. Hung, S.-M. Tseng and C. Chen, “Symbiotic System Models for Efficient IOT System Design and Test,” in Proc. IEEE International Test Conference in Asia (ITC-Asia), Sept. 2017 (to appear).
[10] A. Nordrum, "The internet of fewer things," IEEE Spectrum, pp. 12–13, Oct. 2016.
[11] H.-W. Hung, Peer-Assisted Repair Mechanism for Reliable and Cost-Effective IOT Systems, MS Thesis, National Tsing Hua University, Hsinchu, Taiwan, Jun. 2017.
[12] B. Schroeder, E. Pinheiro, and W.-D. Weber, “DRAM Errors in the Wild: A Large-Scale Field Study,” in Proc. SIGMETRICS ’09, pp. 193-204, 2009.
[13] J. Meza, Q. Wu, S. Kumar and O. Mutlu, “A large-Scale Study of Flash Memory Failures in the Field,” in Proc. SIGMETRICS ‘15, vol. 43, no. 1, pp. 177-190, Jun. 2015.
[14] J. Meza, Q. Wu, S. Kumar, O. Mutlu, “Revisiting Memory Errors in Large-Scale Production Data Centers: Analysis and Modeling of New Trends from the Field,” in Proc. IEEE/IFIP Int. Conf. Dependable Systems and Networks (DSN), 2015.

簡易檢索 / 詳目顯示

相關論文