Management Information Systems, 13TH ED.
MANAGING THE DIGITAL FIRM
Kenneth C. Laudon ● Jane P. Laudon
continued
Learning Track 6: Software Vulnerability and Reliability
In addition to disasters, viruses, and security breaches, defective software and data also pose a
constant threat to information systems, causing untold losses in productivity. An undiscovered
TABLE 1-1 Top Software Glitches 2012
Software glitch Description
The NASDAQ stock exchange The NASDAQ stock exchange computers failed in the first half hour of trading in
Facebook’s IPO stock offering. Investor orders were not recorded so investors
did not know what price they paid for the stock, or even if they owned it. The
BUGS AND DEFECTS
A major problem with software is the presence of hidden bugs or program code defects. Studies
have shown that it is virtually impossible to eliminate all bugs from large programs. e main
source of bugs is the complexity of decision-making code. Even a relatively small program of
Chapter 8: Securing Information Systems
Chapter 8 Learning Track 6 2
continued
60 percent of errors discovered during testing are a result of specifications in the design documen-
tation that were missing, ambiguous, in error, or in conffict.
Zero defects, a goal of the total quality management movement, cannot be achieved in larger
programs. Complete testing simply is not possible. Fully testing programs that contain thousands
THE MAINTENANCE NIGHTMARE
Another reason that systems are unreliable is because computer software traditionally has been a
nightmare to maintain. Maintenance, the process of modifying a system in production use, is the
most expensive phase of the systems development process. In most organizations nearly half of
information systems staff time is spent maintaining existing systems.
Why are maintenance costs so high? One major reason is organizational change. e firm
may experience large internal changes in structure or leadership, or change may come from its
Chapter 8 Learning Track 6 3
continued
FIGURE 8-1 The Cost of Errors Over the Systems Development Cycle.
RESOURCE ALLOCATION DURING SYSTEMS DEVELOPMENT
Views on resource allocation during systems development have changed significantly over the
years. Resource allocation determines the way the costs, time, and personnel are assigned to differ–
ent phases of the project. In earlier times, developers focused on programming, with only about 1
percent of the time and costs of a project being devoted to systems analysis (determining specifi-
SOFTWARE METRICS
Software metrics can play a vital role in increasing system quality. Software metrics are objective
assessments of the system in the form of quantified measurements. Ongoing use of metrics allows
Chapter 8 Learning Track 6 4
be processed in a specified unit of time, online response time, the number of payroll checks printed
per hour, and the number of known bugs per hundred lines of code.
TESTING
e stages of testing required to put an information system into operation are program testing,
system testing, and acceptance testing. Early, regular, and thorough testing will contribute signifi-
cantly to system quality. In general, software testing is often misunderstood. Many view testing as
a way to prove the correctness of work they have done. In fact, we know that all sizable software is
riddled with errors, and we must test to uncover these errors.
Testing begins at the design phase. Because no coding exists yet, the test normally used is a walk–
through—a review of a specification or design document by a small group of people carefully
Electronic commerce and electronic business applications introduce new levels of complexity for
testing to ensure high-quality performance and functionality. Behind each large Web site, such as
Amazon.com, eBay, or E*TRADE, are hundreds of servers, thousands of miles of network cable,
and hundreds of software programs, creating numerous points of vulnerability. ese Web sites
must be built and tested to make sure that they can withstand expected—and unexpected—spikes
Chapter 8 Learning Track 6 5
extremely risky because such problems often stem from the fundamental workings of the system.
To minimize the chance of discovering major structural problems late in the system’s development