Processing time-constrained aggregate queries in CASE-DB

Abstract
In this paper, we present an algorithm to strictly control the time to process an estimator for an aggregate relational query. The algorithm implemented in a prototype database management system, called CASE-DB, iteratively samples from input relations, and evaluates the associated estimator until the time quota expires. In order to estimate the time cost of a query, CASE-DB uses adaptive time cost formulas. The formulas are adaptive in that the parameters of the formulas can be adjusted at runtime to better fit the characteristics of a query. To control the use of time quota, CASE-DB adopts the one-at-a-time-interval time control strategy to make a tradeoff between the risks of overspending and the overhead, finally, experimental evaluation of the methodology is presented.

This publication has 16 references indexed in Scilit: