12.4 Installation


Installation is the process of moving from the current system to the new or enhanced system. This is also refer to conversion activity; converting an old system to a new system. At this stage, all users must give up their reliance on the current system and begin to rely on a new system. There are four approaches of installation; direct, parallel, single-location and phased installation. Which approaches will be used is depends on the organizations and the system's scope and complexity.

12.4.1 Direct Installation

Direct installation is also referring to abrupt cut-over installation. Direct installation is whereby the old system is turned off and the new system turned on Figure 12-2. It's plan as on this date at this time, the old system is terminated and the new system will take over. This approach is a high risk approach because there may still be major problems that won't be uncovered until the system has been in operation. An errors resulting from the new system will have a direct impact on the users and how they do their jobs is depends on the new system. If the system fails to be used on that specific date, delay may occur until the errors are solved. This approach may be necessary when for instance, there is a business policy becomes effective on a specific date and the system can't be implemented before the date. If the organization plan to use this approach, it's important to make sure that everything is success with the system, so that less risky. Although it's known as risky approach, but it may reduce the cost of installation.

12.4 Installation


Figure 12-2: Direct Installation.


12.4.2 Parallel Installation

Parallel installation is when both system; old and new system are operating at the same time until the management decide that the old system can be turned off as in Figure 12-3. This approach is known as riskless approach because all major problems have been solved before the old system is turned off. This approach is an opposite approach with direct installation. Although it's riskless, but it required high costs since the organization needs to run two systems at the same time. Running two similar systems at one time can be very expensive not only in paying the employee salaries, but also in terms of maintenance. Sometimes, users are also confused since they must deal with two systems.


12.4 Installation


Figure 12-3: Parallel Installation.


12.4.3 Single-location Installation

Single location also refers to location or pilot installation. This approach is where the organization is trying out to use a new system at one location and use this experience in deciding how the entire system should be deployed throughout the organization as in Figure 12-4. The one site installation is either use direct installation or parallel installation. It's a middle-of-the-road approach compared with direct and parallel installation. Single location can be referring as branch office, or one department. As soon as the location approved the system, it will be install to the other location. At this stage, the installation at other site can be direct installation since all the major errors have been fixed at the first test

12.4 Installation



location. The advantage of using this strategy is it can minimize the potential damage and costs because the successful is determined at the first site. But the disadvantage is, it may take a longer period to install the whole system.

Figure 12-4: Single-location Installation.

12.4 Installation


12.4.4 Phase Installation

Figure 12-5: Phased installation.


Installation may also require a system acceptance test plan. System acceptance test is a final opportunity where end users, management and information system operation decide either to use or reject the system. System acceptance test is a test performed on the final system wherein users conduct verification, validation and audit tests

12.4 Installation


(Bentley et. al., 2007). It covers three levels of acceptance testing:

Verification testing
Verification testing also refers to alpha testing. This testing runs the system in a simulated environment using simulated data. It's looking for errors and omissions regarding end users and design specifications that were specified in the earlier phase of testing.

Validation testing
Validation testing also refers to beta testing. This testing runs the system in a live environment using real data. A number of items tested during this testing such as:

system performance
peak workload processing performance
human engineering test
methods and procedures test
backup and recovery testing

Audit testing
Audit testing is a test performed to ensure that the system is free from any errors and ready to be placed at the real location.

12.4 Installation