多云协同存储架构及方法研究
发布时间:2018-02-26 18:26
本文关键词: 云存储 多云协同 中间件系统 自适应 分发 出处:《西安电子科技大学》2014年硕士论文 论文类型:学位论文
【摘要】:云存储因其使用便利、性价比高、扩展性强而得到广泛使用。由于失去对物理存储介质的控制,云存储在可靠性、效率以及安全方面都面临严峻的挑战。用户数据完全由云服务提供商控制和管理导致数据丢失、损坏等可靠性问题和数据的安全保护问题;用户网络环境的变化导致数据存取的效率无法达到最优化。针对单一的云存储服务的可靠性问题以及对云备份开销较大的问题,提出了基于数据分片的多云协同存储架构,设计了多云存储中间件系统。该存储架构能够利用现有的多种云存储平台进行协同存储,具有可扩展性、高效性和低开销的优点。在所提出的多云存储架构的基础上,综合考虑多云存储资源和网络服务质量,提出了用户数据自适应分发及整合算法。分析与验证表明相比于Hadoop、Swift等使用的数据分发及整合算法,该算法能够有效提高数据存取的效率。针对多云协同存取存在的数据安全保护问题,提出了数据存取与数据管理、用户访问相分离的信息安全保障机制,给出了基于元数据服务器的数据分片存储和密钥管理方法,克服了云存储数据泄露的问题。基于以上技术,设计和开发了多云协同存储系统,验证所提出的技术。
[Abstract]:Cloud storage is widely used because of its convenient use, high cost performance and strong expansibility. User data is controlled and managed by cloud service provider, which leads to data loss, damage and data security. The efficiency of data access can not be optimized due to the change of user network environment. Aiming at the reliability problem of a single cloud storage service and the large overhead of cloud backup, a multi-cloud cooperative storage architecture based on data slicing is proposed. A multi-cloud storage middleware system is designed. The storage architecture can make use of the existing multi-cloud storage platform for collaborative storage, which has the advantages of scalability, high efficiency and low overhead. On the basis of the proposed multi-cloud storage architecture, Considering multi-cloud storage resources and network quality of service, an adaptive data distribution and integration algorithm is proposed, which is compared with Hadoop Swift data distribution and integration algorithm. The algorithm can effectively improve the efficiency of data access. Aiming at the problem of data security protection existing in cloud cooperative access, the information security mechanism of data access, data management and user access is proposed. This paper presents a metadata server based data partitioning storage and key management method, which overcomes the problem of cloud storage data leakage. Based on the above technology, a multi-cloud cooperative storage system is designed and developed to verify the proposed technology.
【学位授予单位】:西安电子科技大学
【学位级别】:硕士
【学位授予年份】:2014
【分类号】:TP333
【参考文献】
相关期刊论文 前2条
1 赵迪;孙海龙;郑礼全;;基于网络存储虚拟化技术的海量GIS数据存储方法研究[J];国土资源信息化;2008年01期
2 郝树魁;;Hadoop HDFS和MapReduce架构浅析[J];邮电设计技术;2012年07期
,本文编号:1539147
本文链接:https://www.wllwen.com/kejilunwen/jisuanjikexuelunwen/1539147.html