Name of the participant: Marcus Ritter
Description of the IT research project: Cloud technologies have a crucial role to play in digitalisation, Industry 4.0 and IOT. Thanks to the possibility of dynamically allocating resources and adapting services according to demand, they enable great cost and energy savings on the one hand, and a higher number of users and service quality on the other. However, this potential can only be fully exploited if the systems are optimally utilised. Yet this goal is usually contrary to compliance with the Service Level Agreements (SLAs) agreed with the customers, an agreement on service quality and the resources to be made available. In order to avoid bottlenecks and thus a violation of the SLAs, as well as to achieve optimal system efficiency at the same time, accurate models are needed to predict the system performance and the performance of individual cloud services. This research project will develop a new automated methodology to analyse the performance, efficiency and scalability of cloud services and systems.
With the help of the new methodology, a more optimal utilisation of a cloud system can be achieved and thus a reduction of the operating costs for the system infrastructure. At the same time, the energy efficiency of the systems is increased and sustainability is enhanced. Furthermore, the approach enables an optimisation of customer satisfaction, as scalability bottlenecks of a cloud service can be detected and avoided at an early stage, so that an SLA is not violated.
Since previous approaches show deficits in terms of generality, automation and accuracy, the fully automated analysis of the scalability, performance and efficiency of cloud systems and their individual services will be investigated within the scope of this project. Specifically, the following will be investigated:
- The identification of the performance behaviour and the load limits of a cloud service in relation to its scalability
- An analysis of the costs and efficiency of a cloud service in relation to the defined SLAs
- An automated reporting system to convert the findings of the system analysis into recommendations for action, e.g. for system administrators
Software Campus partners: TU Darmstadt, Huawei
Implementation period: 01.07.2021 – 30.06.2023