Performance of Query Optimization Heuristics in Object-Oriented Databases
Kyung-Chang Kim · Database Systems for Advanced Applications · 1991
This paper investigates the performance of query optimization heuristics in object-oriented databases. Query optimization involves the following two stages : (1) Reduce the search space of access plans for a query to be considered during query evaluation using optimization heuristics (2) Estimate the cost of each access plan in (1) and select the cheapest access plan. To compute the cost of each access plan, a cost model is formulated which uses information such as the order of evaluation of classes in a query graph, and the access paths and retrieval algorithms used during the evaluation. The cheapest access plan selected is then used to evaluate the given query. Our performance results show that the optimization heuristics developed are effective and reasonable.