Modeling the Influence of data amount on the response time of CRUD Operations under Oracle

Abstract

In his paper we investigate the influence of data amounts on the response time of CRUD (Create, Read, Update, and Delete) Operations under Oracle 12c. Several sample tables have been designed and implemented in order to investigate the influence of every data amount component (records, attributes and data size) on the response time of CRUD operations. The experimental results show clearly that the response time of CRUD operations depends on the number of records, attributes or the data size. When the number of records, attributes or the data size increases, the response time increases linearly with different average slopes which depend on the type of operation. The highest slope was observed for the UPDATE operation, while the response time for the SELECT operation shows the lowest one. The results show that the response time versus the number of records is increased by 59.349, 5.735, 79.373 and 73.449 seconds for every one million records processed by INSERT, SELECT, UPDATE and DELETE operations respectively. The response time versus the number of attributes is increased by 13.923, 2.436, 49.168 and 23.351 seconds for every attribute added to the sample table and processed by INSERT, SELECT, UPDATE and DELETE operations respectively. The response time versus the data size is increased by 0.632, 0.0029, 0.657 and 0.636 seconds for every megabyte added to the sample table and processed by INSERT, SELECT, UPDATE and DELETE operations respectively. This approach of modeling the response time in a linear form significantly brings out the model of Oracle performance on the response time prediction, which is an important and useful standard for the design and implementation of database applications.