内存数据索引：以处理器为核心的性能优化技术

In-memory index:Performance enhancement techniques leveraging on processors

随着单机内存容量的持续上升,内存数据库技术逐渐取代传统磁盘数据库为数据管理提供更快速的支持.本文分析了设计内存索引结构所需要考虑的基本要素;对目前的内存索引结构进行了分类总结,并分析各结构的优缺点;针对当前应用发展趋势,指出内存索引未来发展的机遇与挑战;最后介绍了我们正在研发的分布式集群感知内存数据库(CLAIMS)中的内存索引结构.

As main memory capacities grows larger and larger, the memory era has arrived andin-memory databases have taken the place of traditional disk-based databases to provide efficientdata management. In this paper, we analyzed the fundamental elements in in-memory indexdesigning： summarized and evaluated the existing index structures, pointing out the futureopportunities and challenges based on the development trend of current applications. Finally, weintroduced our on-going distributed in-memory index studies on the Cluster Aware In-MemorySystem （CLAIMS）.