Curio: A Novel Solution for Efficient Storage and Indexing in Data Warehouses
|
|
Anindya Datta |
Krithi Ramamritham |
Helen Thomas |
|
Muninn Technologies* |
Univ. of Mass., Amherst |
Muninn Technologies |
|
|
Georgia Tech. |
IIT Bombay |
Georgia Tech. |
|
| adatta@cc.gatech.edu | helen@cc.gatech.edu |
Abstract
| The most common approach used to improve On-Line Analytical Processing (OLAP) query performance is to utilize indexes or access structures to quickly access the base data. A major drawback to this approach is that it often incurs significant overhead as the access structures must be stored in addition to the base data. In this demonstration, we present Curio, a data repository and OLAP query server, which provides drastically improved performance for ad-hoc queries, while simultaneously reducing the storage costs associated with warehousing. Curio, recently developed by Muninn Technologies, LLC, is based on a novel paradigm that allows fast access to data, yet does not require indexes. We briefly describe the underlying positional indexing techniques behind Curio and report results from a performance comparison with several leading commercial relational warehousing products in terms of query performance. We demonstrate the improved query performance of Curio over a specific product, namely Oracle. |
| * now Chutney Systems |