【摘要】：隨著移動計算的出現(xiàn)、計算機硬件技術(shù)的發(fā)展與人們?nèi)找嬖鲩L的應(yīng)用需求，越來越多的個人擁有多個智能計算設(shè)備。人們每天的工作與生活都依賴于這些計算設(shè)備上的大量信息數(shù)據(jù)，因此安全有效地共享與管理多設(shè)備間的文件顯得非常重要。 遠程文件同步服務(wù)能夠自動維護多個設(shè)備之間的文件一致性，云存儲技術(shù)與NoSQL數(shù)據(jù)庫技術(shù)為用戶文件數(shù)據(jù)提供了安全、可擴展的備份存儲。本文針對海量信息時代下的個人用戶多設(shè)備之間文件共享與管理的應(yīng)用需求，提出了一種云存儲環(huán)境中的遠程文件同步模型——SyncCS模型。本文重點研究了遠程文件同步模型的系統(tǒng)架構(gòu)、同步協(xié)議以及后臺存儲環(huán)境，完成的主要工作包括： 1．將遠程文件同步技術(shù)與云存儲技術(shù)相結(jié)合，并考慮到傳統(tǒng)點對點文件同步模型的不足，提出了一種C/S架構(gòu)的云存儲環(huán)境中的遠程文件同步模型——SyncCS模型，并構(gòu)建了模型的三層體系結(jié)構(gòu)。 2．設(shè)計了一種適用于SyncCS模型的兩階段文件同步協(xié)議，將文件同步過程分為非實時雙向同步和實時單向同步兩個階段，能夠兼顧文件雙向同步與實時同步兩種不同的用戶需求。同時充分研究并討論了在同步過程中可能發(fā)生的所有文件沖突情況，制定了與兩階段同步協(xié)議相適應(yīng)的文件沖突檢測與處理機制。 3．結(jié)合當(dāng)前炙手可熱的云存儲與NoSQL數(shù)據(jù)庫技術(shù)，研究并構(gòu)造了SyncCS模型后臺云存儲環(huán)境，根據(jù)SyncCS模型中實際的數(shù)據(jù)特點，分別提出了相適應(yīng)的元數(shù)據(jù)與文件數(shù)據(jù)存儲方案。 4．針對SyncCS模型中冗余日志帶來冗余同步操作的問題，提出了一種客戶端日志預(yù)處理優(yōu)化方案；針對大量并發(fā)訪問可能導(dǎo)致元數(shù)據(jù)服務(wù)器效率低下的問題，提出了一種基于Hbase的元數(shù)據(jù)存儲優(yōu)化方案。并通過實驗驗證了兩個優(yōu)化方案的有效性，，從而分別從客戶端與服務(wù)器端對SyncCS模型進行了性能改進。
[Abstract]:With the emergence of mobile computing, the development of computer hardware technology and the increasing demand for applications, more and more individuals have multiple intelligent computing devices. The daily work and life of people depend on a large amount of information data on these computing devices, so it is very important to share and manage the files between multiple devices safely and effectively. Remote file synchronization service can automatically maintain file consistency between multiple devices. Cloud storage technology and NoSQL database technology provide secure and extensible backup storage for user file data. In this paper, a remote file synchronization model in cloud storage environment, SyncCS model, is proposed to meet the application requirements of file sharing and management between individual users and multi-devices in the age of mass information. This paper focuses on the system architecture, synchronization protocol and background storage environment of the remote file synchronization model. The main works are as follows: 1. Combine remote file synchronization technology with cloud storage technology. Considering the shortcomings of the traditional point-to-point file synchronization model, a remote file synchronization model called SyncCS model in cloud storage environment based on C / S architecture is proposed, and the three-tier architecture of the model is constructed. 2. A two-stage file synchronization protocol suitable for SyncCS model is designed. The file synchronization process is divided into two stages: non-real-time bidirectional synchronization and real-time one-way synchronization, which can take account of two different user requirements: bidirectional file synchronization and real-time file synchronization. At the same time, all possible file conflicts in the synchronization process are studied and discussed, and a file conflict detection and handling mechanism suitable to the two-stage synchronization protocol is developed. 3. Combined with the current hot cloud storage and NoSQL database technology, the background cloud storage environment of SyncCS model is studied and constructed. According to the actual data characteristics of SyncCS model, the corresponding metadata and file data storage schemes are put forward respectively. 4. Aiming at the problem of redundant log synchronization in SyncCS model, a preprocessing optimization scheme for client log is proposed. Aiming at the problem that a large number of concurrent access may lead to the inefficiency of metadata server, a metadata storage optimization scheme based on Hbase is proposed. The effectiveness of the two optimization schemes is verified by experiments, and the performance of the SyncCS model is improved from the client side and the server side respectively.
【學(xué)位授予單位】：上海大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2013
【分類號】：TP333

【參考文獻】

相關(guān)期刊論文 前3條

1 程瑩;張云勇;房秉毅;徐雷;;云計算時代的數(shù)據(jù)庫研究[J];電信技術(shù);2011年01期

2 林子雨;賴永炫;林琛;謝怡;鄒權(quán);;云數(shù)據(jù)庫研究[J];軟件學(xué)報;2012年05期

3 周可;王樺;李春花;;云存儲技術(shù)及其應(yīng)用[J];中興通訊技術(shù);2010年04期

本文編號：2356551