摘要: 在内存数据库系统中,针对处理器缓存对提高内存数据库的性能有重要影响的情况,在B+树的基础上提出一种新的缓存敏感的索引——HT索引。将Hash方法和树结构相结合,构造一种适用于内存数据库的索引。结果证明,该索引结构能提高处理器缓存的利用率,其整体操作性能优于传统的缓存敏感索引。
关键词:
索引,
缓存敏感的索引,
B+树,
HT树
Abstract: In memory databases, making good use of the processor caches is of great significance to enhance the performance of memory database. In order to deal with the problem above, based on B+ tree, a cache conscious index called “Hash & Tree Tree(HT Tree)” is proposed. This index, which combines Hash with tree, is constructed for memory database. Result shows that the index can improve the utilization rate of processor caches. It can gain better performance than conventional cache conscious indices.
Key words:
index,
cache conscious index,
B+ tree,
Hash &,
Tree Tree(HT Tree)
中图分类号:
肖富平;罗 军. HT树:缓存敏感的内存数据库索引[J]. 计算机工程, 2009, 35(16): 68-70.
XIAO Fu-ping; LUO Jun. Hash & Tree Tree: Cache Conscious Index of Memory Database[J]. Computer Engineering, 2009, 35(16): 68-70.