Use simulation to explore the effects of database schemas on the performance of database systems

碩士 === 中華大學 === 資訊管理學系(所) === 98 === The main purpose of this study is to develop a simulation methodology for evaluating the performance of the database systems under different database schemas. The simulation methodology is organized as seven steps including (1) simulation scope determination, (2)...

Full description

Bibliographic Details
Main Authors: Hung-Hsiu Tsai, 蔡宏修
Other Authors: Chi-Chung Lee
Format: Others
Language:zh-TW
Published: 2010
Online Access:http://ndltd.ncl.edu.tw/handle/41745519442046849595
Description
Summary:碩士 === 中華大學 === 資訊管理學系(所) === 98 === The main purpose of this study is to develop a simulation methodology for evaluating the performance of the database systems under different database schemas. The simulation methodology is organized as seven steps including (1) simulation scope determination, (2) database schema selection, (3) data import and transformation, (4)simulation environment construction, (5)simulation program coding, (6) simulation program execution, and (7)simulation output analysis. This study further conducts a case study that uses the proposed simulation methodology and the real data of the company to investigate the performance discrepancy among three schemas, namely, horizontal schema, vertical schema, and hybrid schema. The performance evaluation followed the steps of the proposed simulation methodology is made in the case study. In the case study, this study first obtains the database from the shop flow system of the company, and then transfers the horizontal schema database to the vertical schema database and the hybrid schema database. Third, this study finds the queries and the transactions for the daily operations of the company and executes them on the simulation. Finally, this study evaluates the performance of these three schemas in terms of the number of logical reads, the number of physical reads, the number of physical writes and elapsed time. The results show that the hybrid schema has the best performance when compared to the horizontal schema and the vertical schema.