...

Maximizing Use of Academic Computing Resources

by DR. JULIO C. RIVERA, Assistant Professor DR. SANJAY K. SINGH, Assistant Professor DR. FRANK M. MESSINA, Assistant Professor and DR. KHRIS McALISTER, Associate Professor University of Alabama at Birmingham Birmingham, Ala. Current developments in the personal computer and telecommunications arenas make it essential for academic institutions to re-examine their use of computing resources. There now exist many technologies that can enrich student and faculty experiences in the academic environment. In many cases, however, it is essential to tie computing resources into networks to achieve this goal. In this paper, we first examine the rationales for adopting a LAN in an educational environment and point out the pitfalls. Then, based on our personal experiences in designing, implementing and maintaining a 100-workstation LAN, we provide a brief primer on hardware and software requirements for a network. Information Resources Many different resources suited to the needs of students and researchers are available for those who possess the means to exploit them. Generally, these information sources are accessible via telecommunication connections to outside agents who make such resources publicly available. Although many educational institutions participate in these sharing arrangements, resources are not always easily available to students and faculty. Judicious development of academic computer networks can make these resources accessible to everyone. Accessing resources often takes the form of Internet connection, which is now a worldwide network spanning educational, governmental and commercial institutions.1 It is likely the Internet will be expanded further through active government participation (currently being debated). Several types of resources are available on the Internet, ranging from public or private institutions that allow access to information they hold, to a variety of ad hoc interest groups and news groups that discuss topics of interest to everyone.1 Internet users can typically access library catalogs, information servers and electronic bulletin boards scattered throughout the world. New technologies may soon allow access to collections at libraries, enabling the connected educational institutions to share one universe of information sources. While there is some cost to an academic institution for participating on the Internet, it is not normally passed on to users, making the Internet an attractive resource for faculty and students. Table 1 offers a brief list of resources available through the Internet. Interpersonal Communication Connecting computers into a network gives users the ability to communicate with each other. A common problem in many academic settings is the difficulty that students have in locating and exchanging messages with faculty members. The use of electronic mail through a LAN allows two-way communication by routing messages through the network and storing them until the intended receiver is available. In addition, files containing assignments or other material can be sent through most networks. Research into the use of e-mail indicates there is great interpersonal communication benefits to be gained.2 E-mail users often engage in much freer discussions than they would if they were face to face with the other participants.3 Thus, e-mail use can be a source of new ideas and interesting points of view on myriad issues. Of course, it is also possible to misuse e-mail and engage in offensive behavior; users may perceive little fallout from hurling invectives at other participants due to a perceived anonymity. Another big advantage is the ability to send messages directed at groups of users or all users on a system. Many employ this technique to spread news of particular events quickly and also to ask for help, ideas or opinions. This capability provides both an up-to-date source of information and also facilitates dissemination of important events (e.g., Tianammen Massacre). Most of these features are also available on a global basis when a LAN is linked to outside resources such as the Internet. Researchers can take advantage of network connections to share their findings. Beyond messages and file transfer, many forums exist for discussing scientific and other topics. Forums cover everything from the frivolous to the obscure, but are highly productive in terms of information and idea exchange. A prime example of forums in action occurred during the investigation of cold fusion. Interested parties swapped information, asked questions and volunteered opinions, greatly speeding the investigative process to see if the "breakthrough" was real (it was not). The only requirements for joining a forum is access to the Internet and an urge to participate. Teaching Tool LANs can also be used as a teaching tool. In an academic environment, a network can provide faculty with the means to distribute information, assignments, exams and advice. If properly configured, as in an electronic classroom, these networks can also be used to show students how to do certain tasks. Dissemination of assignments and other class information allows participants to break temporal and situational links, which often impede effective communication between faculty and students.4 This makes possible the development of a "virtual classroom," greatly extending educational opportunities to potential students. When used appropriately, such facilities can increase faculty effectiveness and extend the reach of academic programs. The opportunity to enhance teamwork is also provided. Beyond work and information exchange, new software built around network facilities serves to help team participants work better together. For example, Group Support System (GSS) software relies on a network environment to bring together team members in brainstorming sessions, build a team consensus and promote more focused attention to problems.5 Further benefits are available by facilitating group interaction on classwork requiring teamwork and allowing teams to exchange information.6 GSS systems are gaining acceptance in the business environment and have benefits in any activity requiring teamwork to solve complex problems. Managing Computer Resources Another benefit of adopting computer networks is the simplification of computer resource management. Most academic institutions operate computer labs where students have access to software required to complete class assignments. With stand-alone computers, managing these facilities is often difficult, time consuming and inefficient. LANs provide the means to control the propagation of software throughout computing facilities and allow the sharing of expensive resources not generally available to everybody. This creates a uniform computing environment for students and faculty, and reduces the time it takes to learn the use of the available computing resources. An issue of special concern to all institutions is the handling of software licensing requirements. Typically, the situation involves many different pieces of software, each having a different set of requirements regarding how to maintain a legal license. Due to the legal ramifications of not adhering to licensing requirements, it is advisable to manage software distribution centrally through a network. Networks also offer an easy and quick way of configuring software uniformly. Storing software in a central location allows it to be configured appropriately and negates the efforts of "inquisitive" users to alter its configuration. Uniformity also benefits users in that it presents a consistent interface to those with little experience. Finally, should the need arise to change the configuration of any software for workstations, it is relatively easy to do so: simply store the new configuration centrally and propagate it throughout the network. This also reduces workstation requirements for disk storage. A separate but equally important issue in the management of computer resources is the decision of which resources to acquire and how to allocate them. In a stand-alone environment, it is difficult to share expensive resources such as color laser printers, fax modems, scanners, etc.

11059-14616



This usually results in duplication or making some resources unavailable to most users. LANs, however, allow resources to be shared with little or no disruption to users' work. Funds for acquiring devices can be pooled to obtain a fewer number of peripherals, but with each having a more extensive set of features. For example, one could replace several low-quality printers with a single high-quality color laser printer available to all on the network. Similarly, specifications for workstations on a network can be pared down by eliminating hard disk storage. Most network hardware manufacturers allow the loading of all workstation software directly from network server(s). This removes the need for disk storage on workstations (except for diskettes); reduces acquisition costs; and over the life of the workstation, eliminates a source of potential problems. Although not all networks offer this feature, it can help offset the cost of connecting workstations to a LAN. Network Adoption Caveats Along with the tremendous benefits network adoption brings, a commensurate number of problems may occur. Connection to other computers and systems can be an invitation to the machinations of hackers and other "inquisitive" individuals. Even if there is no sensitive data to protect, allowing unrestricted access to network resources may result in these resources being corrupted or misused. It is therefore essential to identify areas of potential vulnerability and adopt appropriate security measures. E-mail is a valuable communications resource that can also be misused or abused. Although messages may not be anonymously sent, lowered inhibitions can lead to harassing or hateful messages directed at individuals. Further, e-mail messages are not guaranteed to be private; they may be read by persons other than the intended recipient. Appropriate measures must be adopted to prevent these transgressions. The situations mentioned above (and others) point out the need for careful thought and management of any network implementation. This often translates into employing personnel specifically responsible for network operation. For most installations this is a full-time job for one or more persons. Although it is preferable to hire someone, it is not absolutely necessary. Many sites can be managed with part-time help, often students. A need for permanent staff becomes evident as the sophistication of the installed network and/or the number of users increase. Hiring staff to manage computing resources d'es not absolve users of decisions on how to best use resources, but it d'es facilitate implementation of those decisions. LAN Hardware Requirements Three types of hardware are usually necessary to implement a network. First, every computer requires a Network Interface Card (NIC) to connect and communicate with the network. Second, the physical connection medium for the network must be installed. Finally, most networks require use of a variety of devices to route and forward messages to intended recipients. At first glance, this is an imposing array of equipment, however, the incremental expense may be relatively small. Network Interface Cards are readily available at very attractive prices, especially in a volume purchase. Also, some computer manufacturers now offer built-in networking capabilities; others are following suit to remain competitive. To physically interconnect computers in a network, choice abound. These range from various types of cables to several forms of wireless connections. The simplest method is the use of existing telephone cable, known as twisted pair. This is the preferred alternative if there is a Private Branch Exchange (PBX) already in place and it has extra pairs of wire not utilized by existing telephones. Even if there is no existing wiring, twisted pair cable is inexpensive and relatively easy to install. Other alternatives are thin or thick coaxial cable and wireless schemes. Thin coaxial cable is inexpensive but is limited to 30 workstations on a network segment. Thick coaxial cable is low cost in itself, but requires the additional expense of transceiver taps to connect to every workstation. Wireless networks, which use radio waves or infrared signals, are higher priced than wired connections. (Fiber is not generally used for smaller LANs, although it is the choice for "backbones" that link multiple LANs.) Finally, depending on the choice of wiring, additional network routing and forwarding devices may be needed. Twisted pair wiring, for example, requires hubs or concentrators to connect all workstations together. These are an extra expense, but also offer advantages when troubleshooting and configuring a network. Thick coaxial cable requires transceiver taps, but allows signals to be carried for much longer distances. Finally, separate networks are often interconnected to extend their reach, which requires bridging and routing devices. LAN Software Requirements The other essential part of any network implementation is the software necessary to interface with the network, manage it and allow user access. This network operating system (NOS) software is loaded on every station in addition to the traditional operating system. It is important to choose a NOS that fulfills the requirements for which the network is being implemented. The particular NOS required varies according to the specific network. Novell NetWare, Microsoft LAN Manager, Artisoft LANtastic, DEC PathWorks, and Banyon VINES are all examples of NOS software. A NOS can control a LAN under either a dedicated server scheme or as a peer-to-peer network. Selecting one depends on the capabilities desired. Generally, server-based networks store both software and any data needed by users on the server. On peer-to-peer networks, users equally share their computers and the peripherals attached to them. Peer-to-peer networks usually do not have as broad a range of capabilities as those with servers (although that difference is getting smaller), but their cost is often substantially lower. Another attractive feature of peer-to-peer is the ability to turn any workstation into a server. Conclusions LANs are a relatively low-cost method of obtaining access to many computing resources, enriching the educational and research experience in the process. Properly configured networks allow users access to remote computing resources, electronic mail, shared peripherals, and shared software and data. Users on these systems can break the temporal and spatial links that bind them to classrooms, meetings, telephone calls and appointments. This liberty gives users greater flexibility with which to pursue their educational objectives and also allows them to participate in a greater community. In addition, from an administrative perspective, important benefits are gained by connecting all computing resources together and centrally managing them. Although networks offer much, thought must be given to the consequences of their adoption. Network security must be addressed, based on both intended use and the potential for misuse. Implementing a LAN requires expenditure of funds along with a need for expertise in its installation and management. However, the potential gains far outweigh these costs. The barriers to adopting and using networks continue to decrease as new technologies spread. Local area networks do not solve all of the problems found in managing computers in an academic environment, but they do help to minimize them while providing new opportunities for academic development. Julio Rivera is an assistant professor in the Department of Management. E-mail: busf029@uabdpo.dpo.uab.edu Sanjay Singh, an assistant professor in the Department of Management, has published in Decision Support Systems and Journal of Informatics and Telematics. E-mail: busf044@uabdpo.dpo.uab.edu Frank Messina, an assistant professor of Accounting and a CPA, has published in The Natural Resources Tax Review. E-mail: busf047@uabdpo.dpo.uab.edu Khris McAlister, an associate professor and chair of the Department of Management at the University of Alabama at Birmingham, has published more than 20 articles and professional papers and is a founding member and past president of the Society for Information Systems, Alabama Chapter. E-mail: busf009@uabdpo.dpo.uab.edu References: 1. Falk, B., The Internet Roadmap, Alameda, CA: Sybex Inc. (1994). 2. Daft, R.L., Lengel, R.H. and Trevino, L.K., "Message Equivocality, Media Selection and Manager Performance: Implications for Information Systems," MIS Quarterly, Vol. 11, No. 3 (1987), pp. 355-366. 3. Hiltz, S.R., Johnson K. and Turoff, M., "Experiments in Group Decision Making -- Communication Process and Outcome in Face-to-Face Versus Computerized Conference," Human Communication Research, Vol. 13, No. 2 (1986), pp. 225-252. 4. Daft, R.L. and Lengel, R.H., "Organizational Information Requirements, Media Richness, and Structural Design," Management Science," Vol. 32, No. 5 (1986), pp. 554-571. 5. Pinsonneault, A. and Kraemer, K., "The Impact of Technology Support on Groups: An Assessment of the Empirical Research," Decision Support Systems, Vol. 5 (1989), pp. 197-216. 6. Watson, R., "A Study of Group Decision Support Systems' Use in Three- and Four-Person Groups for a Preference Allocation Decision," unpublished PhD. dissertation: University of Minnesota(1987). 7. Yanoff, S., "Special Internet Connections," (1993). Compilation of Internet connections available from csd4.csd.uwm.edu.

This article originally appeared in the 05/01/1994 issue of THE Journal.

comments powered by Disqus

Whitepapers