A Semantic Approach for Analyzing the Impact on Service Availability in Commercial Distributed Systems
A critical aspect of distributed systems, especially in the business sector, is the availability of provided services. Due to their decentralized architecture, it is not trivial to predict concrete implications of a malfunction or partial failure of a system component on the availability of individual services in large-scale distributed systems.
Solutions exist for measuring service availability, visually analyzing complex service networks in a manageable graph-based representation, monitoring individual system components, or modeling failure propagation to predict impacts on service availability. However, a scalable solution that unites these partial aspects does not yet exist.
The objective of this thesis is to develop a platform that provides users with an accurate representation of the service network and effective means to monitor and manage its service availability. For this purpose, a requirements analysis must first be carried out and the state of the art in the field of service availability management must be explored. The existing solution approaches are to be classified and evaluated according to the requirements. Finally, a semantic approach for modelling service networks will be designed, implemented and evaluated in order to represent complex relationships between individual system components and to predict potential impacts of malfunctions or failure of single or multiple components on service availability.