Sitemap Contact Us
When: Back to Calendar April 26, 2013 @ 10:00 am - 12:00 pm
Where: YR-459
COSC Conference Room
Contact: Karne, Ramesh
rkarne@towson.edu

Student’s Name:  Uzo Okafor
Department/Program:  Computer and Information Science

Chair, Dissertation Committee:   Dr. Ramesh Karne
Committee Chair phone:    410 704 3955            
email:     rkarne@towson.edu

Title of Dissertation: Transforming SQLITE to run on a Bare PC

Abstract:

This dissertation extends on-going Bare Machine Computing (BMC) research at Towson University.  BMC applications run on a bare machine without any commercial operating system, kernel or other centralized support and are in need of a bare machine database management system.  This research deals with transforming SQLITE DBMS system to run on a bare PC. The SQLITE DBMS is a lean database that runs on Windows or Linux operating system. It is a commonly used as a standalone database across many academic institutions and also a free download. When a DBMS runs on top of an operating system, it uses the operating system’s system calls to access hardware resources. This dissertation will eliminate such system calls and allow the SQLITE DBMS to directly access and manage hardware resources. This novel concept eventually will pave the way to transform other systems and application programs.

 The transformation process poses many daunting challenges and issues. There are variety ways to explore the transformation process.  However, this dissertation proposes a novel approach in transformation methodology using existing tools. It uses Microsoft Visual Studio to develop, test, validate and debug bare PC applications; this step is referred to as pseudo transformation process.  This pseudo transformed code is then used in further transformation process where rest of the OS related dependencies are eliminated. This fully transformed code is now ready to run on a bare PC.

 The primary objective of this research assumes minimal understanding or modification of the SQLITE code during its transformation. This dissertation demonstrates and validates this hypothesis successfully.  After transformation, the functionality and results are validated with its original SQLITE model that runs on a Windows operating system. This investigation serves as a cornerstone for future operating system based application transformation to run on a bare PC or a bare machine. It also lays a foundation to build an automated tool to replace the manual process outlined in this research.

Date of Defense:   April 26 2013
Time:   10 am
Location:    YR-459, 4th Floor COSC Conference Room