近年來關於分散式儲存的相關應用越來越多，像是資料庫、伺服器等級的作業系統以及雲端儲存都有使用分散式儲存的概念，為了有效利用分散式儲存系統中的傳統硬碟，必須要有一個好的資料分配模式，有許多的研究對這資料分配提出了許多不同的見解，但隨著各大企業爭相投入這塊市場，使得分散式儲存的相關應用使用者日漸增多，對分散式儲存系統存取的資料量越來越大，造成儲存系統雖然有好的資料分配方式但因為受限於傳統硬碟的硬體限制，使得效能無法再往上提升。因此為了提升分散式儲存系統的效能，我們將固態硬碟用來取代傳統硬碟，並且根據固態硬碟的特性以及固態硬碟在寫入上的觀察去設計了一個資料分配的模式，我們考慮了寫入資料的資訊以及儲存設備的資訊，有效的以整體速度下降最低的方式去分配資料，最後在實驗中我們將能看出此方法能有效的減少總花費時間。

Distributed storage systems products have become increasingly popular in recent years, and cloud storage is among the most commonly used products. The effective use of traditional hard drives within cloud storage systems necessitates a good data distribution model. Many studies have provided differing views regarding how this data distribution should be conducted and developed. However, with major companies vying to enter this market, the number of cloud storage users and the quantity of data accessed for cloud storage systems has increased. Consequently, although currently equipped with good data distribution methods, the performance of storage systems cannot be further upgraded because of the hardware limitations of traditional hard drives. Therefore, to enhance the performance of cloud storage systems, we replaced traditional hard drives with solid-state drives (SSDs). Additionally, according to the characteristics and data writing methods of SSDs, we designed data distribution models. We considered written data and storage block information, and effectively distributed data with a minimal decline in overall speed. Subsequently, through an experiment, we explored how this technology could successfully reduce total use time.

[1] Gap-Joo Na ; Sang-Won Lee, “A Hybrid Flash Memory SSD Scheme for Enterprise Database Applications”, IEEE Web Conference (APWEB),2010

[2] Jen-Wei Hsieh ; Tei-Wei Kuo ; Li-Pin Chang, “Efficient identification of hot data for flash memory storage systems”, ACM Transactions on Storage (TOS), 2006

[3] Yongqing Zhu ; Yang Yu ; Wilson Yonghong Wang ; Sophia Sufui Tan ; Tee Chong Low, “A Balanced Allocation Strategy for File Assignment in Parallel I/O Systems”, IEEE Networking, Architecture and Storage (NAS), 2010

[4] Xiaojian Wu ; A. L. Narasimha Reddy, “Exploiting Concurrency to Improve Latency and throughput in a Hybrid Storage System”, IEEE Modeling, Analysis & Simulation of Computer and Telecommunication Systems (MASCOTS), 2010

[5] “Managing storage space in a flash and disk hybrid storage system”, Xiaojian Wu ; A. L. Narasimha Reddy, IEEE Modeling, Analysis & Simulation of Computer and Telecommunication Systems(MASCOTS), 2009

[6] lung Sik Park ; Hi-Seok Kim ; Ki-Seok Chung ; Tae Hee Han, “PRAM and NAND flash hybrid architecture based on hot data detection”, IEEE Mechanical and Electronics Engineering (ICMEE), 2010

[7] Soojun Im ; Dongkun Shin, “Storage architecture and software support for SLC/MLC combined flash memory”, IEEE Symposium on Applied Computing(SAC), 2009

[8] Huang-Te Hsu ; Ying-Wen Bai, “Using NAND Flash memory to improve the performance of HDDs”, IEEE Electrical and Computer Engineering (CCECE), 2010

[9] Hyun-Seob Lee ; Hyun-Sik Yun ; Dong-Ho Lee, “HFTL: hybrid flash translation layer based on hot data identification for flash memory”, IEEE Consumer Electronics, 2009

[10] Jung, S. ; Lee, Y. ; Song, Y.H., “A process-aware hot/cold identification scheme for flash memory storage systems”, IEEE Consumer Electronics, 2010

[11] Dongchul Park ; Du, D.H.C., “Hot data identification for flash-based storage systems using multiple bloom filters”, IEEE Mass Storage Systems and Technologies (MSST), 2011

[12] Li-Pin Chang ; Tei-Wei Kuo, “An adaptive striping architecture for flash memory storage systems of embedded systems”, Real-Time and Embedded Technology and Applications Symposium, 2002

[13] Gong Zhang ; Chiu, L. ; Ling Liu, “Adaptive Data Migration in Multi-tiered Storage Based Cloud Environment”, IEEE Cloud Computing (CLOUD), 2010

[14] Klonatos, Y. ; Makatos, T. ; Marazakis, M. ; Flouris, M.D. ; Bilas, A., “Azor: Using Two-Level Block Selection to Improve SSD-Based I/O Caches”, IEEE Networking, Architecture and Storage (NAS), 2011

[15] Qingsong Wei ; Veeravalli, B. ; Bozhao Gong ; Lingfang Zeng ; Dan Feng, “CDRM: A Cost-Effective Dynamic Replication Management Scheme for Cloud Storage Cluster”, IEEE Cluster Computing (CLUSTER), 2010

[16] Kyuwoon Kim ; Sanghyuk Jung ; Yong Ho Song, “Compression ratio based hot/cold data identification for flash memory”, IEEE Consumer Electronics (ICCE), 2011

[17] Tin-Yu Wu ; Wei-Tsong Lee ; Yu-San Lin ; Yih-Sin Lin ; Hung-Lin Chan ; Jhih-Siang Huang, Computing, “Dynamic load balancing mechanism based on cloud storage”, Communications and Applications Conference (ComComAp), 2012

[18] Yongkun Wang ; Goda, K. ; Nakano, M. ; Kitsuregawa, M., “Early Experience and Evaluation of File Systems on SSD with Database Applications”, IEEE Networking, Architecture and Storage (NAS), 2010

[19] IOTTA website: http://iotta.snia.org/

[20] Dushyanth Narayanan ; Austin Donnelly ; Antony Rowstron, “Write off-loading: Practical power management for enterprise storage”, ACM Transactions on Storage (TOS), 2008

[21] Iometer website: http://www.iometer.org/