RA3: Distributed Information Systems Architectures and Web Services Computing

- How IT (Technology components: protocols, etc.) innovation affects IS (Systems: e-delivery, etc.) capabilities -

Overview

"A Web service is viewed as an abstract notion that must be implemented by a concrete agent. The agent is the concrete entity (a piece of software) that sends and receives messages, while the service is the abstract set of functionality that is provided [see Figure below] (W3C 2003, 7).

Figure: Service Oriented Model of Web Services Architecture (W3C 2003, 33)

"Ever since the first program was divided into modules, software systems have had architectures, and programmers have been responsible for the interaction among the modules and the global properties for the assemblage" (Shaw and Garlan 1996, xii).

Introduction

In many organizations, the information technology (IT) infrastructure, with its many different systems provided by numerous vendors, has emerged over many years--even decades--often without a strategic plan. With a more disruptive business environment on the one hand, and, therefore, sharply increased requirements for integration and flexibility, and more distributed information systems (IS) choices, on the other, companies often overlook, however, urgently need an IS architecture strategy.

Evolution of IS Architecture Styles
Computing styles or IS architectures have evolved from highly centralized systems with mainframes to decentralized and distributed systems with workstations and PCs. The principal enablers of this evolution of computing styles have been the rapid increases in both processing power ("Moore's law") and the bandwidth of connections between computers.

In the 1990s client/server (C/S) computing has evolved as the dominant distributed systems architecture paradigm (Ryan and Andersen Consulting 1996). Client/server computing is commonly defined as "communication between computing processes that are classified as service consumer (clients) and service providers (servers)" (Umar 1997, 11; Turban et al. 2001, 538-541; Hoffer et al. 2002, 532-538). In other words, any component that submits a request becomes a client, and any component that replies becomes a server. Hence, the C/S style is also referred to as request/reply architecture. Furthermore, the distinction between client and server is a conceptual one only--the same computer can be a client in one transaction and a server in another interaction.

Web Services: "The Next Big Thing"?
Over the past decade many new distributed systems architectures have emerged, such as the common object request broker architecture (CORBA; www.omg.org) and Sun Microsystem's J2EE (Java Enterprise Architecture). The latest innovation is Web services (WS) technology. It is presented by IT vendors, systems integrators, and many trade journals as "The Next Big Thing" in computing (The Economist 2002). While Internet technology such as the TCP/IP protocols have improved interconnectivity, WS technology promises improved application-level interoperability. This could increase sharing of functionality and from an economic perspective, unit cost of automation could be greatly reduced, which could boost the extent of overall business process automation in the economy.

Key Issues, Research Objectives and Questions

In the past, many firms have complemented or replaced mainframe computing with client/server computing. However, choosing client/server as the default IS architecture can result in a mismatch between the client/server system's request/reply style and the pattern of interaction (mutual or reciprocal action between parties) inherent in a firm's business models. This pattern of interaction includes the message passing, workflow, and distribution of tasks among parties in a business system.

The extent of mismatch has increased lately due to the emergence of disruptive concepts such as peer-to-peer computing and Web services technology. The mismatch can cause high operating cost, bandwidth bottlenecks, inflexibility, and, in a worst-case scenario, an outright lack of scalability.

While business models and systems architectures are well described, respectively, in the management and computer science literature, surprisingly, very little attention has been given to linking them appropriately. Therefore, my goal is to clarify choices. I pose the following research questions:

How should a business align an IS architecture with a business model and strategy?

How does Web services technology affect IS architecture design? How does it affect IS capabilities--the competencies, skills, and tacit knowledge that an organization develops to effectively acquire, deploy, and leverage its IT investments in pursuit of its business strategy? Does Web services technology moderate the relationship between systems flexibility and and tightness of enterprise integration? In order to increase efficiency many companies have invested in enterprise resource planning (ERP) and customer relationship management (CRM) systems only to find out that tight integration has reduced their flexibility-the ability to quickly respond to a new sales opportunity or to add new product features, for example. Can Web services technology solve this dilemma?

To what extent is there dissonance between the 'hype'--anticipated, hypothesized Web services-related business effects--and the status quo? As of Spring 2003, with many Web services components still under construction, the future of Web services remains uncertain. Most analysis remains focused on hypothesized and anticipated business effects (Iyer et al. 2003, Patil & Saigal 2001, Hagel & Brown 2001). How does Web services technology affect business models--particulalry outsourcing--and the functional decomposition of business processes? Are enterprise systems indeed becoming more modular or agent-based?

Approach

My approach is twofold: I complement theory development with case study based research. Please refer to Projects and Publications to see how we are making efforts to provide answers to these questions.

Acknowledgment

Research into Web Services is supported by a grant from .

Chris Langdon also gratefully acknowledges thoughtful discussions about Web services-based prototypes and case studies with Scott M. Rose, Global Managing Director of Development, Accenture Technology Labs, Chicago.

References

The Economist. 2002. Timely Technology-New Kinds of Software Could Make Companies Both More Integrated and Flexible. February 2: 5-7.

Hagel, J., and Brown, J. S. 2001. Your Next IT Strategy. Harvard Business Review (October): 105-113.

Hoffer, J. A., J. F. George, and J. S. Valacich. 2002. Modern Systems Analysis and Design. Prentice Hall: Upper Saddle River, NJ.

Iyer, B., J. Freedman, M. Gaynor, and G. Wyner. 2003. Web Services: Enabling Dynamic Business Networks. Communications of the AIS 11: 525-554.

Patil S., and Saigal, S. 2001. When Computers Learn to Talk: A Web Services Primer. The McKinsey Quarterly [electronic document] (accessed 01/31/02); available from http://www.mckinseyquarterly.com; Internet.

Ryan, H. W., and Andersen Consulting. 1996. Practical Guide to Client-Server Computing. Auerbach: Boston, MA.

Shaw, M., and D. Garlan. 1996. Software Architectures--Perspectives of an emerging discipline. Prentices Hall: Upper Saddle River, NJ.

Turban, E., E. McLean, and J. Wetherbe. 2001. Information Technology for Management--Making Connections for Strategic Advantage. 2nd ed., John Wiley & Sons: New York, NY.

Umar, A. 1997. Object-Oriented Client/Server Internet Environments: The Modern IT Infrastructure. Prentice-Hall: New York, NY.

World Wide Web Consortium. 2003. Web Services Architecture. Technical Report [electronic document] (accessed 10/8/03); available at http://www.w3.org/TR/2003/WD-ws-arch-20030808/.

(Created by: csl, 02/14/03; last modified by: csl, 10/26/03.)