Tutorial 4:

 

SRE of Web Site Construction

Speaker: Professor Norman F. Schneidewind

 

[ About the Speaker ]  [ Abstract ]  [ Outline

 

 

About the Speaker:

Dr. Norman F. Schneidewind is Professor of Information Sciences and Director of the Software Metrics Research Center in the Division of Computer and Information Sciences and Operations at the Naval Postgraduate School, where he teaches and performs research in software engineering and computer networks. Dr. Schneidewind is a Fellow of the IEEE, elected in 1992 for "contributions to software measurement models in reliability and metrics, and for leadership in advancing the field of software maintenance". He is the developer of the Schneidewind software reliability model that is used by NASA to assist in the prediction of software reliability of the Space Shuttle, by the Naval Surface Warfare Center for Trident software reliability prediction, and by the Marine Corps Tactical Systems Support Activity for distributed system software reliability assessment and prediction. This model is one of the models recommended by the American National Standards Institute and the American Institute of Aeronautics and Astronautics Recommended Practice for Software Reliability. In addition, the model is implemented in the Statistical Modeling and Estimation of Reliability Functions for Software (SMERFS), software reliability-modeling tool. He has published widely in the fields of software reliability and metrics.

In 1993 and 1999 he received an award for outstanding research achievements by the Naval Postgraduate School. He was Chairman of the Working Group that produced the IEEE Standard 1061-1992, Standard for a Software Quality Metrics Methodology and its revision in 1998. In 1993 he was given the IEEE Computer Society's Outstanding Contribution Award "for work leading to the establishment of IEEE Standard 1061-1992". In addition, he was given the IEEE Computer Society Meritorious Service Award "for his long‑term committed work in advancing the cause of software engineering standards". Prof. Schneidewind was recognized by the IEEE Standards Board in the "Awards Spotlight" of the IEEE Standards Bearer, Vol. 13, No., February 1999, p.6, for the publication of his standard: "IEEE Standard for a Software Quality Methodology (Revision), 1998". Richard J. Holleman, Chairman of the IEEE Standard Board also recognized Prof. Schneidewind's excellence in standards development in his letter of 4 June 1999. The IEEE Software Engineering Standards Committee has appointed Prof. Schneidewind as the Chair of the Working Group to develop a Standard Dictionary of Measures of the Software Aspects of Dependability and a Guide for the Use of Standard Dictionary of Measures of Software Aspects of Dependability.

He is a member of the IEEE\CS Fellows Committee that evaluates candidates for that grade and has been elected three times a member of the Management Board of the IEEE Software Engineering Standards Committee that sets policy for the development of IEEE software engineering standards. He serves as Associate Editor, IEEE Transactions on Software Engineering He was recognized for his contributions to the IEEE Computer Society by being named to the "Golden Core" of volunteers.

Dr. Schneidewind served for several years as a member of the Department of Defense Software Engineering Institute Measurement Steering Committee. Currently he is a member of the DoD Software Collaborators Group. He also served on the AIAA Software Reliability Blue Ribbon Committee that produced the ANSI/AIAA Recommended Practice on Software Reliability. Dr. Schneidewind is also a former Editor of the IEEE Computer Magazine Standards Department and Vice Chair for Standards of the IEEE Technical Council on Software Engineering for several terms. He is past Chair, IEEE Technical Committee on Simulation. He has served on the National Science Foundation Software Engineering and Languages panel to review research proposals in this area.

Dr. Schneidewind organized the first Conference on Software Maintenance in 1983. He was the General Chair for the International Conference on Software Maintenance in 1996 and in 2000. He was also the General Chair of the International Symposium for Software Reliability Engineering-1994 and Program Chair of the 9th International Conference on Distributed Computing Systems,1989. Prof. Schneidewind has been appointed to the International Symposium on Software Reliability Engineering policy committee that has the responsibility of drafting a long-range plan for the content and organization of the symposium.

Dr. Schneidewind is an IEEE Computer Society Distinguished Visitor Program speaker and an IEEE Computer Society Chapters Tutorial Program speaker and has been an invited speaker at the University of California, Berkeley; University of California, Irvine; Linkoping University, Sweden; Technical University of Dresden, Germany; speaker in the "Distinguished Speakers Series" at Florida Atlantic University; The National University of Singapore; British Telcom Laboratories, U.K.;  Laboratoire d' Automatique et d'Analyse des Systems, France; NASA Johnson Space Center; Lawrence Livermore National Laboratory; Sandia National Laboratories; Santa Clara Valley IEEE Product Safety Technical Committee and System Safety Society; Santa Clara Valley Software Quality Association; Dallas-Fort Worth Chapter of the IEEE Reliability Society; and the Naval Surface Warfare Center. He has also been a keynote speaker at the Stan Ackerman Institute, Technical University of Eindhoven, Eindhoven, The Netherlands; Reliable Software Technologies - Ada-Europe '96, Switzerland; The Institute of Decision Science, Claremont McKenna College; and the Conference on Software Maintenance, Miami, Florida. He has also presented many tutorials at International Symposium for Software Reliability Engineering, International Conference on Software Maintenance, Quality Week, and Software Technology Conference.

Prior to joining the Naval Postgraduate School, Dr. Schneidewind held several technical management positions in the computer industry, where he directed a number of projects, including the National Marine Data Program, Executive Office of the President; National Traffic Data Center, U.S. Department of Transportation; studies on information systems for the Library of Congress; Bay Area Rapid Transit System computer control system; California State Utilities Commission; Navy Submarine Logistics project, and the Marin County data processing study. He also managed the development of a 100 million record title company tax and lien application and was the manager of a 60-person computer center that processed these applications. He received the Commissioner's Award of the U.S. Bureau of Customs for "Contributions to the Data Processing Advisory Panel".

Dr. Schneidewind has a B.S.E.E., University of California (Berkeley); a M.S.E.E. and a M.S.C.S., San Jose State University; a M.S.O.R. (ENGR) and a Doctorate with a major in O.R., University of Southern California. He is a member of Eta Kappa Nu and Tau Beta Pi engineering honor societies and Sigma Xi research society and holds the Certificate in Data Processing (CDP) from the Institute for Certification of Computer Professionals. He is listed in “Who’s Who in Science and Engineering”.

 

 

Abstract:

Web site construction is one of the most important activities in today’s Internet economy. While a great deal has been written about implementing Web sites, very little has been said about the factors of reliability, availability, maintainability, usability, accessibility, performance, and security, and the tradeoffs that must be made among these factors. For example, accessibility and security are in direct conflict, as are performance and security; we could build a Web site with maximum security and zero accessibility! In discussing these factors and tradeoffs, we must consider both the server side and the client side and the interaction between the two: any feature or function implemented at the Web server will affect the usability and functionality as perceived by the client. If, for example, we choose to provide many graphics at our Web site, a price will be paid in reliability and performance, as seen by the user. Another interesting aspect is how we manage to provide continuous access to the Web site when changes must be made to the site. Additionally, there has been overemphasis about the performance of Web servers, at the expense of what the user, as a client, experiences. 

Web site applications have unique characteristics that set them apart from traditional applications. Thus, new thinking and SRE models must be developed for Web sites compared to standalone or even local network models. A major factor that stimulates this need is that a public Web site is exposed to the world community for access, unlike most applications where the user community and its requirements can be defined. Therefore, the needs and accessibility requirements of a much larger user set must be considered in developing the site. Related to this dynamic is that the resources of the user in terms of browser version, monitor size, screen resolution, etc. may be vastly different from and unknown to the developer who designs the site. Thus, although a Web designer may have a certain clientele in mind when implementing a site, the potential clientele and their expectations could be vastly different. 

One of the most interesting examples of this phenomenon is distance learning, wherein a Web site may be designed to provide a learning experience for a defined subject, objectives, and student body, but the actual audience could vary considerably from the intended one because various users surf the Internet looking for various objects including “free” courses. The challenge of providing effective distance learning increases if live video broadcasts of the instructor’s lectures, using a scan of the Web site material on the PC screen, are combined with student access to the Web site from their PCs outside of lecture times. Screen resolution and font size that may be satisfactory for the latter type of access may be unsatisfactory for the former. 

I propose to flesh out these issues and tradeoffs and to provide the attendees with a SRE approach to Web site development, using my practical experience in distance learning Web site development as a framework.

 

 

Outline:

1. Overview of Web Site Construction

a. Obtaining Web site server space and access privileges

b. Using development tools

c. Creating HTML files

d. Testing HTML files on local PC

e. Loading HTML files on server

f. Testing Web site as a client

g. Providing file download and print capabilities

h. Creating external links to Web sites with related material

 

2. Web Site Metrics Definitions

a. reliability

b. availability

c. maintainability

d. usability

e. accessibility

f. performance

g. security

 

3. Web Site Reliability and Usability Factors (examples)

a. Quality of  images: Graphics that are readable on a local PC may not be readable on a Web site.

b. Integrity of URLs: Links that work on a local PC may not work on a Web site.

c. Font size: Fonts that are readable on a local PC may not be readable on a Web site.

d.  Browser version: the client’s browser version may not be compatible with the developer’s version.

e. Display resolution: the client’s screen size and resolution may not be compatible with what the developer used in developing the site.

f. File download and print options: Web sites must allow complete file and print capabilities and not be limited to a single page.

g. Providing adequate Web site resolution when the site is used in conjunction with video and audio broadcast to distance learners: resolution and fonts that may be adequate for Web access from the user’s PC may be inadequate, when using video broadcast. 

 

3. Web Site Maintainability

a.   How to maintain a Web site that must have 24 hour access.

b.   Techniques for checking out modifications on a local PC before committing them to the Web site.

c.   Techniques for editing and maintaining changes to HTML files.

 

4. Web Site Security, Accessibility, and Availability

      a. Placement of firewalls and proxy servers.

1) Effect on client usability and accessibility.

b. Tradeoffs among security, accessibility, and availability            .                      

c. Impact of Ipv6 on reliability, security, accessibility, and availability.

d. Use of Virtual Private Networks to effect security.

 

5. Web Server and Client Metrics

a. Server Side

1) Performance: request rate, network bandwidth.

b. Client Side

1) Performance: request response time.

2) Availability: percentage of time available, percentage of successful logons.

3) Reliability: percentage of successful accesses and presentations of items of interest.

 

6. Distance Learning Application

a. Lessons learned from building Web site.

b. Discussion of implications of reliability, availability, maintainability, usability, accessibility, performance, and security data collected from the Naval Postgraduate School Web site.