A Test Methodology for Mars

Abstract

In this chapter we proceed to describe a test methodology for distributed real-time systems. Of course, it is not possible to give a general test methodology; it depends both on the underlying target system architecture and on the design methodology used for developing the application software. This is because the target system architecture influences to a great extent the testing techniques that may be favorably applied and/or the cost of applying a particular testing technique to the system, as has been extensively discussed in Chapter 3. The particular design methodology, on the other hand, defines the levels of abstraction an application system will undergo during its development and may be viewed at during or after its development — and thus constrains these levels of abstraction for the test methodology. In other words, the design methodology already imposes some elements of Organization (see Chapter 2) onto the software development process, and it is clearly counterproductive if the test methodology violates or rejects the Organization already in place. Further, the design methodology is likely to be influenced by the target system architecture as well.