Objectives: Introduction Over View of System Analysis and Design
Download 0.94 Mb. View original pdf
|
- Navigate this page:
- 8.2.3Error Detection and Correction
- 8.2.4 Error Tolerance
| 8.2.2 Error Avoidance There are three approaches to reliability namely, error avoidance, error detection and error tolerance. Under error avoidance, developers and programmers make every attempt to prevent errors from occurring at all. The emphasis on early and careful identification of user requirements in another way this objective is pursued. Analysts must assume that it is impossible to fully achieve this objective. Errors will occur despite the best efforts of very competent people. 8.2.3Error Detection and Correction This method uses design features that detect errors and make necessary changes to correct either the error while the program is in use or the effect on the user, so that a failure does not occur. Correcting user errors, such as misspelling keywords or entering invalid commands, is one remedy. Error detection in programs is handled in a similar manner. For example, a program that calculates the productivity of a waiter or waitress in a restaurant by dividing the total revenue from meals served into the hours worked should not fail when employees do not serve anything. When a blinding snowstorm prevents customers from coming to a restaurant, employees will accumulate working time but will not have sales. The program should detect the divide – by – zero error and correct for it in order to keep the system running properly. Unfortunately, many programs fail when a situation like this one occurs. Even though it may not happen for several years after the system is installed, the error is therefrom the day of development. The failure occurs later. 8.2.4 Error Tolerance Error tolerance strategies keep the system running even in the presence of errors. The United States National Aeronautics and Space Administration (NASA, for example, designs its systems to be error – tolerant through the use of redundant hardware. In one space program, redundant on – board computers and computer voting are used to process data in parallel, so results can be compared. Two computers process the data on location, course correction and compare the results with those produced by two other computers processing the same data. A fifth computer is available to break a tie should one occur. If needed, a sixth computer stored away in an accessible storage compartment can quickly replace one of the other computers that has been damaged or failed. Another manner of error tolerance is the use of degraded processing. With this strategy, the user receives less service than the system was designed to provide, but that is considered abetter alternative in some cases than having no service at all. For example, many electric power generation and distribution facilities in North America are computer – controlled. Suppose that on a record – breaking hot day the system becomes overloaded and the computer control center is unable to correctly process allocation data and keep up with the power demands. Rather than risk damaging the power distribution network, the computer automatically shuts down part of the network. By providing degraded service, the computer tolerates a software error without failing. Download 0.94 Mb. Share with your friends:
|