Reducing the braking distance of an SQL query engine

Reducing the braking distance of an SQL query engine (1998)

by Michael J. Carey , Donald Kossmann

Venue:	In Proc. of the 24th VLDB Conf
Citations:	84 - 6 self

Datum	Value	Source
TITLE	Reducing the braking distance of an SQL query engine	INFERENCE
AUTHOR NAME	Michael J. Carey	SVM HeaderParse 0.2
AUTHOR AFFIL	IBM Almaden Research Center	SVM HeaderParse 0.2
AUTHOR ADDR	San Jose, CA 95120	SVM HeaderParse 0.2
AUTHOR NAME	Donald Kossmann	SVM HeaderParse 0.2
AUTHOR AFFIL	University of Passau	SVM HeaderParse 0.2
AUTHOR ADDR	94030 Passau, Germany	SVM HeaderParse 0.2
ABSTRACT	In a recent paper, we proposed adding a STOP AFTER clause to SQL to permit the cardinality of a query result to be explicitly limited by query writers and query tools. We demonstrated the usefulness of having this clause, showed how to extend a traditional cost-based query optimizer to accommodate it, and demonstrated via DB2-based simulations that large performance gains are possible when STOP AFTER queries are explicitly supported by the database engine. In this paper, we present several new strategies for efficiently processing STOP AFTER queries. These strategies, based largely on the use of range partitioning techniques, offer significant additional savings for handling STOP AFTER queries that yield sizeable result sets. We describe classes of queries where such savings would indeed arise and present experimental measurements that show the benefits and tradeoffs associated with the new processing strategies. 1	SVM HeaderParse 0.2
YEAR	1998	INFERENCE
VENUE	In Proc. of the 24th VLDB Conf	INFERENCE
VENUE TYPE	CONFERENCE	INFERENCE
PAGES	158--169	INFERENCE
CITATIONS	14 found	ParsCit 1.0