Measuring of Oracle Optimizer Performance for Object-Relational Database Entities

Abstract

SQL Optimization is the process of choosing the most efficient way to execute SQL such as Data manipulation language (Select, Insert, Update, Delete). Relational model represents the object through the use of relations; on the other hand the main aim of ORDBMS model is to represent the special characteristics of OOP. The first model deals with the Relation & Correlated sub queries, while the second one uses collection types (Object type, Varrays, Nested table). Explain plane is used to examine exactly how oracle execute SQL statements " oracle SQL analyze" provides a facility for easily generate any explain plan . DML language (SQL) as tools for any comparison can be achieved by the statistical performance parameter (Elapsed time, CPU time, Logical block read, Physical block read …). Statistical analysis using ANOVA TABLE is used to judge between comparisons. We found that nested table was the most effective type of data, the object type was the least effective (consume lowest time of execution). We also found that increasing data size will increase any of performance data parameter until we reach size of 40% the time begins to decrease. It is found that SELECT, UPDATE and DELETE have the most influence on the performance parameter, respectively.