Summary | Solaris System Management (New Riders Professional Library)

Testing is a discipline that requires a high level of skill to ensure that all areas of a system or application are thoroughly examined and exercised. It is also an aspect of system development that is frequently overlooked or shortened because of budgetary constraints.

The Y2K issue highlighted the need for thorough testing and, because of the immovable deadline, forced businesses to identify and prioritize their requirements. The result of carrying out a detailed inventory meant that many companies ended up in much better shape, with a more streamlined and efficient way of working.

The system manager benefits from a number of testing phases throughout the development and implementation of a new system or application, most notably system testing and performance testing, because of the justification that he can provide for requesting additional funding for the systems.

Coexistence testing and performance testing are important because they verify that the system or application will work with other products, that it fits in with the existing infrastructure, and also that it can satisfy the predicted requirement for concurrent usage, allowing for predicted peaks of demand.

Acceptance testing is often used in larger installations, particularly when a third-party supplier is involved and the cost is much higher. The "user community" gets to test the functionality against the original specification, using it as it will be used in the operational environment. The application or system will not be signed off or paid for in full unless the user acceptance-testing phase completes successfully.

The importance of end-to-end testing is fundamental to any new application or system in that it exercises the entire business function, often incorporating several existing systems and processes. It is an excellent means of building customer confidence in the new implementation.

The final testing phase before implementation allows the system or application to be seen in its live environment, but without the risks or impact of full operational running. When carried out in two stages, minor performance issues can be resolved by the system manager before the full implementation takes place.

Top