The use of pre-computed aggregate data is a common technique to address performance in Data Warehouse systems (DW), but the adoption of aggregates must be transparent. An aggregate retargeting mechanism redirects a query to available aggregate table(s) whenever possible. In this work, we present an aggregate retargeting mechanism that deals with multiple fact table schemas (MFTS), which are composed of many fact tables related by conformed dimensions. The algorithm provides two types of transparency: a) aggregate unawareness, and b) MFTS join complexity unawareness. The algorithm requires simple metadata, and it is non-intrusive. The paper presents the retargeting algorithm, required metadata and performance experiments.
展开▼
