基于一致性树分布的数据分布式存储方法

Distributed data storage method based on consistent tree distribution

随着云计算和大数据技术的发展,传统的单一存储介质的数据存储方式已经不能满足大数据处理的需求,在这样的背景下,分布式数据存储得到了广泛的应用。然而,目前存在的几种分布式存储方式并不能够完美地满足分布系统的需求。为了更有效地实现数据的分布式存储和冗余备份,采用一种新的基于一致性树分布(CTD)的分布式存储方法,并提出基于该方法的备份策略,实现数据索引与存储位置的映射。该方案具有负载平衡、无单点故障问题、扩展性高且易于实现的优点。同时提出了基于一致性二叉树分布(CBTD)的应用方案。通过对应用系统实例的分析,验证该方法能够很好地满足分布式系统的数据平衡、负载均衡和扩展性需求。

With the development of cloud computing and big data, traditional single storage medium based data storage cannot meet the demands of large data processing, thus distributed data storage has been widely used recently. However, some existing distributed ways of storage cannot perfectly meet the needs of the distribution system. In order to achieve the distributed data storage and redundancy more effectively, a new distributed method and backup strategy based on Consistent Tree Distribution （CTD） was proposed to achieve the location mapping of data indexing and storage. The new method featured load balancing, no single point of failure, high scalability and easy programming. In addition, a method based on Consistent Binary Tree Distribution （CBTD） was also put forward. Analysis of the application system verifies that the method can satisfy the data balance, load balance and high scalability requirements of the distributed system.