Histogram-Based Approximation of Set-Valued Query-Answers
VLDB '99 Proceedings of the 25th International Conference on Very Large Data Bases
Improving the Processing of Decision Support Queries: The Case for a DSS Optimizer
IDEAS '01 Proceedings of the International Database Engineering & Applications Symposium
Coarse-grained optimization: techniques for rewriting SQL statement sequences
VLDB '03 Proceedings of the 29th international conference on Very large data bases - Volume 29
CHICAGO: a test and evaluation environment for coarse-grained optimization
VLDB '04 Proceedings of the Thirtieth international conference on Very large data bases - Volume 30
A statistics propagation approach to enable cost-based optimization of statement sequences
ADBIS'07 Proceedings of the 11th East European conference on Advances in databases and information systems
| Hi-index | 0.00 |
Query generators producing sequences of SQL statements are embedded in many applications. As the execution time of such sequences is often far from optimal, their optimization is an important issue. Therefore, in [5] we proposed a rule-based optimization approach, which we called CGO (Coarse-Grained Optimization). Our first prototype used a heuristic, priority-based control strategy to choose the rewrite rules that should be applied to a given statement sequence. This worked well but there is still potential for improvements. Thus, in [4] we have introduced an approach to provide cost estimates for statement sequences which is the basis for a cost-based CGO optimizer. It exploits histogram propagation and the optimizer of the underlying database system for this purpose. In this demonstration, we want to showcase the functionality and the effectiveness of our approach. Thereto, we present a prototype of a cost-estimation component for statement sequences which implements this approach. It includes a graphical user interface to explain the histogram-propagation process and to report the results of the cost-estimation process. In the setup for this demonstration, we use a TPC-H benchmark database with an appropriate set of sequences as sample scenario.