KNOWLEDGE TRANSFORMATION ENVIRONMENT

INDIVIDUAL PROJECT PLAN

Ed Sykes

Professor Computer Science

School of Computing and Information Management

Sheridan Centre for Animation and Emerging Technologies

Sheridan College,

1430 Trafalgar Road,

Oakville, Ontario

Canada

L6H 2L1

Copyright © 2001, Ed Sykes


PHASE: ONE

1.1Literature searches and investigation:

1.1.1Linux kernel:

1.1.1.1The degree to which the operating system supports multi-processor, parallel programming.

1.1.2Linux on various hardware platforms:

1.1.2.1Determining the current status of where Linux is used:

1.1.2.1.1(For instance, Intel processor machines, Macintosh machines, etc.)

1.1.3Linux on N-processor machines

1.1.4Linux and networked computing:

1.1.4.1Distributed computing environments using Linux.

1.1.5Oracle on Linux

1.1.5.1Stability investigation

1.1.5.2Performance investigation 

1.1.5.2.1benchmarks:

1.1.5.2.2Oracle on Linux vs. Oracle on Windows, Oracle on Solaris, etc.

1.1.5.3Investigation of current corporate environments:

1.1.5.3.1hardware 

1.1.5.3.2software 

1.1.5.3.3network topology

1.1.5.3.4business orientation 

1.1.5.3.5Information Technology Department organization:

1.1.5.3.5.1training / skill set / future direction

1.2Refinement of the research design:

1.2.1TBD

1.3Scope of the research:

1.3.1Linux:stability, and performance investigation involving:

1.3.1.1networking computing environments, and

1.3.1.2N-processor machines (2 <= n <= 6 ), mid-scale mini computers

1.3.2Oracle on Linux:

1.3.2.1stability, and performance investigation on:

1.3.2.1.1a)single processor machines

1.3.2.1.2b)N-processor machines

1.3.2.1.3c)networked computing environments



1.4Opportunities for supporting academics:

1.4.1Numerous programs at Sheridan College:

1.4.1.1Enterprise Database Management: 

1.4.1.1.1Operating Systems course, many of the Oracle specific courses throughout this program.

1.4.1.2Computer Science Technology and System Analyst Diploma programs:

1.4.1.2.1Programming courses: (e.g., perl, c, c++, java on Linux)

1.4.1.2.2Networking courses: (e.g., Linux networking capabilities, distributed computing technologies, etc.)

1.4.1.2.3Operating System courses 

1.4.1.2.4Database courses: (e.g., Oracle content within curriculum)

1.5Opportunities for supporting partnerships:

1.5.1Numerous distributions of Linux vendors:

1.5.1.1Debian, Red Hat, Slackware, Caldera, Corel Linux, etc.

1.5.2Oracle Corporation

1.5.2.1Oracle Canada is already supporting Sheridan in two distinct ways:

1.5.2.1.1Sheridan relies on Oracle for it's operation

1.5.2.1.2Oracle Academic Initiative program (OAI):

1.5.2.1.2.1Enterprise Database Management program

1.5.2.1.2.2All programs in SCIM where there is curriculum content referring to databases.

1.5.3IT Departments:

1.5.3.1Any Corporation / Business in which: 

feasibility study is required for:

a)Operating System:selection of / benefits of Linux / pros / cons

b)Networked Computing solutions (distributed computing -- Linux)

c)N-processor machine research findings:

(system stability and performance)

d)database is required for optimal stability and performance to house critical business data.

1.6Evaluation of success:

1.6.1To Be Determined (TBD)



PHASE:TWO

2.1Prototype implementation:

2.1.1Preliminary experimentation of Linux on different hardware platforms:

2.1.1.1e.g., Intel single-processor machines, Macintosh, etc.

2.1.2N-processor machines:

a)acquire access to machine(s) on which to perform experiments.

b)acquire latest stable Linux kernel for N-processor 

c)install / configure Linux kernel for machine

d)design experiments to determine:

stability / performance / benchmarks of N-processor Linux machine

versus 

single-processor machine 

2.1.3Networked Computing (Linux cluster):

a)setup environment for Linux cluster of single processor machines

b)install / configure necessary software for distributed computing on the cluster

c)design experiments to determine:

stability / performance / benchmarks of Linux cluster

versus 

single-processor machine 

2.1.4Oracle and Linux:

2.1.4.1N-processor machines:

a)acquire access to machine(s) on which to perform experiments:

at least 1 N-processor Linux machine

at least 1 single-processor Linux machine

b)install Oracle 

c)configure Oracle for N-processor machine:

(MTS:Multi-Threaded Server configuration parameters)

(parallel server cartridge)

d)design experiments to determine:

stability / performance / benchmarks of N-processor Linux machine

versus 

single-processor machine, both running Oracle.

2.1.4.2Networked Computing (Linux cluster): [Optional: based on feasibility study]

a)configure Linux cluster for Oracle installation

b)install Oracle on Linux cluster

c)design experiments to determine:

stability / performance / benchmarks of Oracle on Linux cluster

versus 

Oracle on single-processor machine 

2.2Perform experiment:

2.2.1To Be Determined (TBD)

2.3Refine methods and processes:

2.3.1To Be Determined (TBD)

2.4Replicate results:

2.4.1To Be Determined (TBD)

2.5Demonstrate incremental successes:

2.5.1To Be Determined (TBD)

PHASE:THREE

3.1Dissemination of knowledge:

3.1.1Publish findings:

3.1.2Present findings:

3.2Professional / Business communities.

PHASE:FOUR

4.1.Investigate new avenues for research:

4.2.Investigate new areas for potential application of research:

4.3.Openings for cross-disciplinary and interdisciplinary research.

4.4.Consider potential areas for incubation of new ideas / potential breakthrough opportunities.



3.2PERSONNEL REQUIREMENTS

i)One full-time or part-time person performing:

50% - 75% research (and the remaining for teaching)

weekly

continuous and consistent over the 5 year period

ii)Two technical assistants who are available to assist / troubleshoot technical problems specific to research for this project.

reliance:on a need by need basis for the researcher

continuous over the 5 year period

3.3TECHNOLOGY ANDSERVICE REQUIREMENTS

New Purchases:

i)N-processor computer (e.g., Dell 6400 4-processor computer or more powerful)

super-user access to N-process machine to install Linux, install Oracle

ii)1 single-processor computer to install Linux and Oracle (base comparison machine for research)

New Purchases:

iii)several single-processor computers to setup/configure Linux cluster 

(minimum of 4 of the exact same machines in ii) above)

iv)super-computer access to perform:

- Linux installation

- Oracle installation

3.4SPACE REQUIREMENTS

Lab Space:

i)Sufficient room to setup a Linux cluster environment.

ii)Sufficient room to setup a N-processor system.

Office Space:

ii)room for research personnel and supporting technician



3.5OPERATIONAL BUDGET

YEAR Budget

To Be Determined (TBD)

3.6TRAVEL BUDGET

To Be Determined (TBD)

3.7CONFERENCE FEES

YEARBudget

To Be Determined (TBD)

3.8MEMBERSHIP FEES (Indicate organization)

To Be Determined (TBD)

3.9PARTNERSHIP PLANNING

3.9.1ACADEMIC PARTNERSHIPS

University of Waterloo

University of Toronto

University of Brock, Western, Lakehead, and Windsor (joint Ph.D. program)

3.9.2BUSINESS AND CORPORATE PARTNERSHIPS

Oracle Corporation

Mercer, Toronto, Ontario

(others pending…)



4.0COLLEGE TO BUSINESS CONTRACT PARTNERSHIP AGREEMENTS

Please indicate any opportunities for contracted research from industry which you feel may support the research program in your area.

Oracle Corporation

Linux Distribution Corporations:

Red Hat, Caldera, Corel, Slackware, and others

5.0PRESENTATION, PUBLISHING AND SCREENINGS

5.1Anticipated papers:

i)Linux Kernel: multi-processor system investigation:configuration, setup and recommendations

ii)Stability Investigation and Results of N-processor Linux computer systems

iii)Performance Investigation and Results of N-processor Linux computer systems

iv)Parallel programming on N-processor Linux computer systems:Part I

v)Parallel programming on N-processor Linux computer systems:Part II

vi)Stability Investigation and Results of a Linux cluster environment

vii)Performance Investigation and Results of a Linux cluster environment

viii)Stability Investigation of ORACLE on an N-processor Linux computer system

ix)Performance Investigation of ORACLE on an N-processor Linux computer system.

x)ORACLE on Linux in a distributed networked computer environment:Feasibility Study.

xi)Linux:N-processor systems:Future directions

xii)Linux:Distributed systems:Future directions

xiii)ORACLE:Future directions

xiv)ORACLE and Linux:Future directions

5.3Conferences and Workshops:

I have presented at the following Conferences:

i)Linux and Oracle:Make IT Happen, May 25, 2000. Algonquin College.

ii) Linux and Oracle:Make IT So!,Academic Conference, May 18, 2000, Sheridan.

In the future, I feel confident that companies, and institutions that organize similar conferences and workshops will be very excited to offer workshops in the area of Oracle and Linux.This would be an opportunity for the material in the research papers above to be presented.

ORACLE has many conferences worldwide in an effort to showcase leading technologies of their products on all supported platforms.ORACLE would be very interested in supporting conference topics in the area specific to ORACLE on Linux.

Educational institutions would be quite interested in the research conducted.Since Unix is often the major operating system in these types of organizations and Oracle being the database vendor, the prospects of direct applicability of this project's finding would be well received.The parallel between Unix and Linux is extremely close.Thus, the research on the Linux operating system would be quite beneficial to personnel in these settings.

Linux conferences are bountiful. The Linux community is entirely worldwide, and there would be ample opportunities to present my findings to this global community in a number of different media and formats.

6.0IMPACT ON FACULTY

Faculty whose interest and expertise is in Oracle and/or Linux would find the papers and presentations informative and useful.The faculty from the following programs would benefit the most:Computer Science Technology, System Analyst, and Enterprise Database Management.

7.0IMPACT ON STUDENTS

Effects on the development of existing programs:

a)Undergraduate programs:

Computer Science Technology and System Analyst diploma programs:

i)Programming courses: (e.g., perl, c, c++, java on Linux)

ii)Networking courses: (e.g., Linux networking capabilities, distributed computing technologies, etc.)

iii)Operating System courses 

iv)Seminars course:Parallel Programming using Linux

v)Database courses

b)Post-Graduate programs:

Enterprise Database Management

Undergraduate opportunities:

The Computer Science Technology program courses:

- Project Proposal course

- Project course

Graduate opportunities:

There would be provisions for students to become directly involved in research projects at this level.

8.0 NEW ACADEMIC PROGRAMS

1)Applied degree investigation, feasibility and formation.As an example, a High Performance Computing program which would extend the student’s programming experience and operating system knowledge to a level of thinking, designing, and developing in a parallel mode, thus utilizing effectively multiple processors in the computer.

2)Workshops can be organized to provide the necessary education and knowledge to personnel in the community who are actively involved in the areas of ORACLE and Linux.Sheridan's SCAET centre is an excellent location to offer these workshops.

9.0ADDITIONAL EXTERNAL GRANTS

External grants would be beneficial to see the success of this project through.

NSERC would be a means to secure the necessary equipment to fund specific components of this project.