Implementation of Eddy scheme in adaptive query using pipeline-join

The object oriented data model is introduced to handle more complex data. Since early optimizers were designed to use with a relatively simple relational data model, new generation of extensible optimizer were developed to meet the requirements of evolving database systems. The query optimization system is assembled by several join relations, the number of joins are evaluated continuously base on the number of inputs and preferred outputs. The Eddy scheme is assembled by pipeline-joins that operated to generate sub-joins that used for search criteria being processed. From the experiment we obtained that the performance of normal-query and optimal-query are better than adaptive-query for searching data under 3000 tuples, however for data over 3000 tuples the performance of adaptive-query is better than normal-query and optimal-query.