Diploma and Master Theses (authored and supervised):
"Trustworthy Measurement and Arbitration of Service Level Agreements in the Cloud";
Supervisor: S. Schulte, M. Borkowski;
Institute of Information Systems Engineering, Distributed Systems Group,
final examination: 2018-02-26.
Service Level Agreements state commitments between service providers and customers, in which quality aspects and constraints are regulated. In the field of computing, for instance cloud computing, they are used to define requirements in terms of resource provisioning and consumption. In order to verify the contract compliance, a real-time monitoring of the agreed quality parameters is required. A fundamental problem in the monitoring of Service Level Agreements is trustworthiness. If the measurement or arbitration is done at one of the two signature parties, i.e., service provider or customer, trust issues of the opposing party arise, since different economic interests are pursued. In the course of this work, a framework is developed which enables a trustworthy monitoring of Service Level Agreements at cloud services. A Trusted Third Party that is employed for the arbitration analyzes measurements from both the provider and the customer to detect possibly occurring contract violations. Technologies like aspect-oriented programming or software agents are used to allow for the extraction of quality parameters without being invasive to the existing software. An implementation of the presented approach as well as an evaluation in a public cloud test environment is carried out. The results show that the monitoring of the framework is accurate and trustworthy, while requirements for the transparency, i.e., low resource consumption and little intrusion, are met.
Created from the Publication Database of the Vienna University of Technology.