In distributed systems applications must compete for access to resources. Existing algorithms for competing use of resources often only offer the best effort-guarantee. Best effort has no guarantees for the time data uses to its goal. It can only guarantee that data reaches its goal. This guarantee isn't enough for some applications. It's for instance very important for multimedia applications that data reaches its goal in time.
A solution to this problem is to negotiate about Quality of Service. This means that the application, the underlying system and the resources agree on QoS-values that the underlying system and the resources can offer and that satisfies the application.
Middleware is a platform for distributed systems. A type of middleware is componentbased. Component architectures shall simplify programming on distributed systems by dividing the functional aspects and the non-functional aspects in the system. Component Architectures consists of independent components that can be plugged together to a composition.
QuA (Quality of Service aware component Architecture) is a QoS-aware component architecture. QuA should be able to monitor the environment and adapt to changes. The goal of this thesis is to develop a monitoring framework for this architecture. The thesis focuses on what kind of information from the environment that is needed to collect, how this information should be collected, what the framework needs to do with the values before they are sent out and how they should be sent out. It also defines what kind of components and roles this framework should have.
To represent the environment the best way the monitoring framework has to find the present state of the environment and also predict future values. The framework should also do calculation on values allready collected to produce more complicated values.