基于商业运营模式的云存储系统结构设计与分析

发布时间：2018-03-17 18:10

本文选题：云存储　切入点：商业运营模式　出处：《华中科技大学》2012年硕士论文　论文类型：学位论文

【摘要】：自因特网流行以来，网络数据数量以前所未有的速度增长着，传统的存储模式已经无法跟上数据增长的脚步。而基于网络存储的云存储(Cloud Storage)是解决数据存储问题的主力军。其发展潜力不仅表现在它那低廉的代价和安全的服务，，更为重要的是它能够让客户只需拥有输入输出终端就能够正常工作。传统云存储的商业运营结构是客户直接与云存储运营商交互。由于访问接口、收费机制、数据存放形式等原因，客户存放在多个云存储运营商上的数据无法相互交互。典型的云存储拓扑结构Master-Slave和P2P都存在各自的不足。为云存储系统设计了一种新的商业运营模式，在该商业模式的基础上分别从功能、拓扑结构为云存储设计了一种结构。该结构能够解决存放在多个存储运营商上的数据相互交互问题，客户可以根据自己的需要去选择安全级别，同时融合Master-Slave和P2P两种拓扑结构的优点。在这个结构基础上，对云存储系统进行抽象分析，运用元胞自动机理论为存储元胞设定状态、状态转换策略，通过系统的自我调节，可以解决负载均衡、热点数据处理以及数据副本控制等问题。通过在CloudSim上部署仿真模型，采用随机访问请求生成机制，分别对只有读请求和读写请求共同存在两种情况进行仿真。使用matlab工具对仿真数据进行分析。仿真结果表明，为存储元胞设置好状态和状态转变规则后，系统将能够靠自我调节来实现负载均衡，同时处理热点数据的发生和数据副本数量的控制。
[Abstract]:Since the popularity of the Internet, the amount of online data has grown at an unprecedented rate. The traditional storage mode has been unable to keep up with the growth of data. Cloud Storage based on Network Storage is the main force to solve the problem of data storage. More importantly, it enables customers to function as long as they have an I / O terminal. The traditional business operation structure of cloud storage is that customers interact directly with cloud storage operators. The data stored by customers on multiple cloud storage operators cannot interact with each other. Typical cloud storage topology Master-Slave and P2P have their own shortcomings. A new business operation mode is designed for cloud storage system. On the basis of the business model, a structure is designed for cloud storage from function and topology, which can solve the problem of data interaction between multiple storage operators. Customers can choose the security level according to their own needs, and combine the advantages of Master-Slave and P2P topology. Based on this structure, the cloud storage system is analyzed abstractly, and the state of storage cell is set by using cellular automata theory. The strategy of state transition can solve the problems of load balancing, hot data processing and data copy control through self-adjustment of the system. By deploying the simulation model on CloudSim and adopting the random access request generation mechanism, we simulate the two situations of only read request and read and write request respectively. The simulation data are analyzed by using matlab tool. The simulation results show that there are only read requests and read and write requests. After setting up state and state transition rules for storage cells, the system will be able to achieve load balancing by self-adjustment, at the same time, it can handle the occurrence of hot data and the control of the number of data replicas.
【学位授予单位】：华中科技大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP333

【参考文献】