Indexing without the Index: Scalable Multidimensional Aggregation for Data Warehouses

Aggregation plays an important role in data warehousing and has received much attention to date. However, existing techniques do not suuciently address the issue of making the computation of multidimensional aggregates scale with increasing dimensionality. At the same time the beneet of taking full advantage of the hard disk geometry is often overlooked. This paper presents the Multiresolution File Scan (MFS) approach which is based on a selection of at les which are accessed with fast sequential I/O operations. Its simple structure and low storage overhead allow MFS to scale to high dimensionality while making the best use of the increasing transfer speed of modern hard disks. We show that MFS out-performs multidimensional index structures, even if these structures are bulk-loaded and hence optimized for query processing. Our approach can incorporate a priori knowledge about the query workload and is applicable to all distributive (e.g., COUNT, SUM, MAX, MIN) and algebraic (e.g., average) aggregate operators .