摘要: 提出一种数据网格环境下的分布式副本定位算法PM-chord,以完成对所需数据副本的快速定位,降低存储和更新开销。PM-chord取逆时针方向为chord环的正方向,对节点和数据编码后按前缀匹配的原则查询数据,同时增加前继副本机制。分析及实验表明,该算法具有很高的副本定位效率,能够有效解决网格中的查询“热点”问题,具有良好的可靠性和可行性。
关键词:
数据网格,
副本,
副本定位
Abstract: Data replication is a general mechanism to improve performance and availability for large-scale data-intensive applications. However, ensuring efficient and fast access to one or more replicas of desired data is a challenging problem. To solve this problem, a Prefix Matching-chord (PM-chord) algorithm which leads into chord protocol in P2P field is proposed. PM-chord takes anti-clockwise as positive, searches data by using prefix matching principle and adds the predecessor replication mechanism. Analysis and experiments show that PM-chord has good performance on replica location, is effective on hot spots question in data grids, and can achieve reliability and adaptability of implementation.
Key words:
data grid,
replica,
replica location
中图分类号:
王福业;高敬阳;危胜军. 基于PM-chord算法的数据网格副本定位方法[J]. 计算机工程, 2009, 35(2): 55-57.
WANG Fu-ye; GAO Jing-yang; WEI Sheng-jun. Replica Location Method Based on PM-chord Algorithm in Data Grids[J]. Computer Engineering, 2009, 35(2): 55-57.