具有保證多虛擬儲存之公平性與多固態硬碟之負載平衡性的架構

簡易檢索 / 詳目顯示

回結果列表

研究生：	戴健宇 JIAN-YU DAI
論文名稱：	具有保證多虛擬儲存之公平性與多固態硬碟之負載平衡性的架構 Virtual Storages with Fairness on Physical SSDs with Load Balance
指導教授：	吳晋賢 Chin-Hsien Wu
口試委員:	吳晋賢 Chin-Hsien Wu 陳雅淑 Ya-Shu Chen 張原豪 Yuan-Hao Chang 張立平 Li-Pin Chang
學位類別：	碩士 Master
系所名稱：	電資學院 - 電子工程系 Department of Electronic and Computer Engineering
論文出版年：	2020
畢業學年度：	108
語文別：	中文
論文頁數：	49
中文關鍵詞：	固態硬碟、公平性、負載平衡性、虛擬化
外文關鍵詞：	SSD, Fairness, Load Balance, Virtualization
相關次數：	點閱：669 下載：5
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

現今固態硬碟（Solid-State Disk, SSD）因體積小、耗能低、抗震動性質高、安靜、高速存取、非揮發性記憶體等優點現已成為市面上手機、桌上型個人電腦、攜帶型個人電腦乃至各種嵌入式系統之儲存裝置的首選，同時現代越來越多人採取多SSD的架構來提升存取效率以及擴充儲存容量，但是也因此引發了一些問題。第一當多人同時存取同一SSD時，由於SSD或是更上層（即OS）中缺乏公平性策略，使得SSD的資源相互分配不均。其二當多SSD在執行時，會因為上層的存取模式（Access Pattern）的關係，導致每一個SSD的負載量不一致的問題。而為了解決上述問題，本論文提出一個新穎的架構與演算法來讓上層多使用者競爭存取多SSD裝置時可以保證多使用者間相互公平，同時在下層多SSD裝置間可以保證其負載為相互平衡。根據我們實驗結果表明，與基於簡易的公平性分配法比較下，平均公平性分數最大可以提升3.1倍，並且平均負載平衡性分數最大可以提升1.5倍。

Nowadays, Solid-state disk (SSD) has become the best choice of storage in smart phones, Desktop Computers, Notebooks, and various embedded systems because of its brilliant advantages compared with Hard Disk Drive (HDD) such as small size, low-power consumption, high resistance of physical shock, silence, quicker access time, lower latency and non-volatile feature. At the same time more and more people adopt a multiple SSDs architecture to improve efficiency and expand storage capacity, but it would cause some problems. First, when simultaneously access the same SSD, since SSDs or the upper layer (i.e. OS) has lack of fairness strategy, so that uneven distribution of shared resource of SSD. Second, when multiple SSDs in the execution, because the upper has different access pattern, resulting in the load of SSD has different from each other. In order to solve the above problem, this paper presents a novel architecture and algorithms to make the multiple users access to shared resource can ensure that fair for each other, also make the multiple SSDs can ensures that load balancing for each other. According to experimental results show that, compare with simple fairness-based scheduling method, our proposed method can improve the average fairness score by up to 3.1x, and improve the average load balance score by up to 1.5x.

中文摘要    I
Abstract    II
目錄    III
圖目錄    V
表目錄    VI
公式目錄    VII
第一章    Introduction    1
第二章　Background and Motivation    3
2.1 NVMe SSD    3
2.2 Motivation    5
2.2.1 Fairness in SSD    5
2.2.2 Load Balance in SSDs    5
2.2.3 Related Work    5
第三章　Virtual Storages with Fairness on Physical SSDs with Load Balance    7
3.1 System Architecture    7
3.2 Virtual Storages with Fairness    8
3.3 Physical SSDs with Load Balance    12
3.4 Coordinator of Fairness and Load Balance    14
3.4.1 Load-Balance-aware Tuning    16
3.4.2 Fairness-aware Tuning    20
3.5 Implementation of Parameters    22
3.5.1 Calculation of IOPS and I/O Size    23
3.5.2 Calculation of Response Time    24
3.5.3 Classifying I/O Characteristic    25
第四章　Experimental Setup and Evaluation    26
4.1 Experimental Setup    26
4.1.1 Trace-based Simulator    26
4.1.2 Workloads    27
4.1.3 Comparison Metrics    28
4.2 Experimental Evaluation    29
4.2.1 Fairness Evaluation    30
4.2.2 Load Balance Evaluation    33
第五章　Conclusion    35
參考文獻    36
附錄    39


                                

[1] 2019. NVM Express 1.4 Specification. https://nvmexpress.org/resources/specifications/.
[2] Arash Tavakkol, Mohammad Sadrosadati, Saugata Ghose, Jeremie Kim, Yixin Luo,Yaohua Wang, Nika Mansouri-Ghiasi, Lois Orosa, Juan Gómez-Luna, and Onur Mutlu. 2018. FLIN: Enabling Fairness and Enhancing Performance in Modern NVMe Solid State Drives. In 45th ACM/IEEE Annual International Symposium on Computer Architecture, ISCA 2018, Los Angeles, CA, USA, June 1-6, 2018. 397–410.
[3] Reetuparna Das, Onur Mutlu, Thomas Moscibroda and Chita R. Das, “Application-Aware Prioritization Mechanisms for On-Chip Networks,” in MICRO, 2009.
[4] Yoongu Kim, Dongsu Han, Onur Mutlu and Mor Harchol-Balter, “ATLAS: A Scalable and High-Performance Scheduling Algorithm for Multiple Memory Controllers,” in HPCA, 2010.
[5] Hans Vandierendonck and Andre Seznec, “Fairness Metrics for Multi-Threaded Processors,” IEEE Computer Architecture Letters, 2011.
[6] Stan Park and Kai Shen. 2012. FIOS: A Fair, Efficient Flash I/O Scheduler. In 10th USENIX Conference on File and Storage Technologies (FAST). 13.
[7] Kai Shen and Stan Park. 2013. FlashFQ: A Fair Queueing I/O Scheduler for Flash-Based SSDs. In USENIX Annual Technical Conference (ATC). 67-78.
[8] Da-Wei Chang, Hsin-Hung Chen, and Wei-Jian Su. 2015. VSSD: Performance Isolation in a Solid-State Drive. ACM Transaction on Design Automation of Elec-tronic Systems, TODAES. 1–33.
[9] Jian Huang, Anirudh Badam, Laura Caulfield, Suman Nath, Sudipta Sengupta, Bikash Sharma, and Moinuddin K. Qureshi. 2017. FlashBlox: Achieving Both Performance Isolation and Uniform Lifetime for Virtualized SSDs. In 15th USENIX Conference on File and Storage Technologies, FAST 2017, Santa Clara, CA, USA, February 27 - March 2, 2017. 375–390.
[10] Jiahao Liu, Fang Wang, and Dan Feng. 2019. CostPI: Cost-Effective Performance Isolation for Shared NVMe SSDs. In48th ACM International Conference on Parallel Processing, ICPP. 25:1–25:10.
[11] Bryan Suk Kim. 2018. Utilitarian Performance Isolation in Shared SSDs. In 10th USENIX Workshop on Hot Topics in Storage and File Systems, HotStorage.
[12] Chunghan Lee, Tatsuo Kumano, Tatsuma Matsuki, Hiroshi Endo, Naoto Fuku-moto, and Mariko Sugawara. 2017. Understanding Storage Traffic Characteristics on Enterprise Virtual Desktop Infrastructure. In Proceedings of the 10th ACM International Systems and Storage Conference, SYSTOR’17, 13:1–13:11.
[13] SNIA IOTTA Repository Home. http://iotta.snia.org/
[14] Dushyanth Narayanan, Austin Donnelly, and Antony Rowstron. Write off-loading: Practical power management for enterprise storage. In Proceedings of 6th USENIX Conference on File and Storage Technologies, FAST’08, San Jose, CA, February 2008. 256–267
[15] Miryeong Kwon, Jie Zhang, Gyuyoung Park, Wonil Choi, David Donofrio, John Shalf, Mahmut T. Kandemir, and Myoungsoo Jung. 2017. TraceTracker: Hard-ware/software co-evaluation for large-scale I/O workload reconstruction. In IEEE International Symposium on Workload Characterization (IISWC). 87–96.
[16] Aayush Gupta, Youngjae Kim, Bhuvan Urgaonkar, “DFTL: a flash translation layer employing demand-based selective caching of page-level address mappings.” In Proc. Of ASPLOS’09, 2009.

簡易檢索 / 詳目顯示

相關論文