关键词: 列存储, 联机分析处理（OLAP）, 多查询优化, 全局计划, 操作复用

Abstract: This paper provides a multi-query optimization strategy to achieve the share and reuse of column scan, column join and aggregation operations in column store OLAP (on-line analytical processing) systems which can easily lead to large I/O and CPU overhead. According to the features of column-stores and OLAP applications, the paper proposes a series of transformation rules to generate a single global query plan for a set of related queries mapped from a certain OLAP require. Aiming at the share and reuse of operations, the paper also introduces four new defined nodes: the filter node, the group by node, the merge node and the aggregation node. At the same time, it makes an improvement to the MuGA (multiply group by algorithm), and uses the filter node, merge node and join node to mark group number for each tuple in dimension and fact tables, to achieve the share of column scan and column join operations. For the aggregation node, the paper proposes a multi-phase aggregation algorithm combined with compound group number of the fact table to implement effective aggregation reuse. The experimental results on the benchmark data sets SSB (star schema benchmark) also verify the effectiveness of the multi-query optimization strategy.

Key words: column store, on-line analytical processing (OLAP), multi-query optimization, global query plan, operation reuse