以整合式資料處理策略提升雲儲存之服務品質

Quality of Service Enhancement by using Integrated Data Manipulation Strategy in the Cloud Storage Environment

陳柏煒(Bo-Wei Chen)

指導教授 ： 王淑卿

朝陽科技大學/資訊學院/資訊管理系/碩士(2013年)

摘要

由於電腦網路頻寬、硬體設備及相關網路應用服務的持續發展與演進，使得需要儲存的資料量不斷的擴增。架構在分散式系統及網際網路的雲端運算(Cloud Computing)是一種具有可靠性、可用性、規模性、與彈性使用的一種新興之服務模式。在現今已有越來越多使用者選擇將檔案存放至雲儲存(Cloud Storage)環境中，若在雲端環境上運行一些和儲存資源緊密結合的應用程序，此時儲存性能對應用程序的整體實現效果有著關鍵性的影響。因此如何提升雲儲存的效能，進而增加雲端服務提供者的服務品質(Quality of Service；QoS)，為本研究探討的主題。
在本研究中提出了整合式資料處理策略，針對了雲儲存環境的檔案放置、資料壓縮、資料比對的速度、以及檔案傳輸等議題進行探討並提出較佳的解決機制。其中，檔案放置機制透過分群的方式放置，能夠使各儲存節點的負載達到負載平衡；在資料壓縮機制中，提出了輕量化資料重複刪除(Data Deduplication)機制，以降低傳統資料重複刪除技術所帶來的額外成本；在資料重複比對中，則針對傳統布隆過濾器(Bloom Filter)的不足進行改良，將其結合索引的功能，使其能夠以極快的速度完成比對動作，其中比對的時間複雜度能達到O(1)；最後，在資料傳輸的議題中，提出了節點競爭傳輸的方法，透過該方法，能夠使檔案傳輸的速度更快，並可獲取穩定之傳輸品質。

關鍵字

布隆過濾器 ； 資料壓縮 ； 資料放置 ； 雲端儲存 ； 資料重複刪除 ； 檔案傳輸 ； 資料索引

並列摘要

Network bandwidth and hardware technology is developing rapidly, resulting in the vigorous development of the Internet. Cloud Computing is a new concept that provide high reliability, availability, scalability, and flexibility service which was constructed in the distributed systems and the Internet. In recent years, more and more user uses the cloud storage and services, so how to promote quality of cloud storage service are the main objectives of this study.
In this study, an Integrated Data Manipulation Strategy is proposed, the topics of data placement, data compression, file index, and data transport are discussed and the related mechanisms are proposed. In the topic of data placement, the concept of interest-group in peer to peer network is used to ensure the load balance between storage nodes. In the topic of data compression, a new data deduplication mechanism is proposed to reduce the extra cost of traditional data deduplication method. In the topic of file index, the traditional Bloom Filter is modified that index characteristics are combined and the complexity of search time can be derived only O(1). The last topic is data transport, a competition mechanism is proposed between transports nodes that can reduce the transmission time and provide better quality of service to user.

並列關鍵字

Cloud Storage ； Bloom Filter ； Data Deduplication ； Data Transport ； File Index ； Data Compression ； Data Placement

參考文獻

[1] G. Antichi, A.D. Pietro, D. Ficara, S. Giordano, F. Russo and F. Vitucci(2010), “Achieving Perfect Hashing through an Improved Construction of Bloom Filters,” Proc. of IEEE International Conference on Communications, pp. 1-5.

[3] B.H. Bloom(1970), “Space/time Trade-offs in Hash Coding with Allowable Errors,” Commu. of the ACM, vol. 13, no. 7, pp. 422–426.

[4] G. Brunette, R. Mogull(2009), “Security Guidance for Critical Areas of Focus in Cloud Computing,” V2.1, Cloud Security Alliance.

[5] R. Buyya, C.S. Yeo and S. Venugopal(2008), “Market-oriented Cloud Computing: Vision, Hype, and Reality for Delivering IT Services as Computing Utilities,” Proc. of 10th IEEE International Conference on High Performance Computing and Communications, pp. 5-13.

[6] T. Chen, F. Liu, and N. Xiao(2009), “RADPA Reliability-aware Data Placement Algorithm for Large-scale Network Storage Systems,” Proc. of the IEEE International Conference on High Performance Computing and Communications, pp. 648-653.