Design and implementation of a service-monitoring module based on elements of self-awareness computing and reinforcement learning algorithms

• Grant number: 20/10288-5
• Support type: Scholarships in Brazil - Scientific Initiation
• Effective date (Start): October 01, 2020
• Effective date (End): September 30, 2021
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal researcher: Frank José Affonso
• Grantee: William Fernandes Dorante
• Home Institution: Instituto de Geociências e Ciências Exatas (IGCE). Universidade Estadual Paulista (UNESP). Campus de Rio Claro. Rio Claro , SP, Brazil

Abstract

Actually, it is noted that our society is increasingly dependent on software systems to perform daily tasks. Thus, it is expected that such systems will also be able to operate under uncertain conditions, without human interruptions or interventions. The major causes of uncertainties range from changes in the operating environment to variations in the objectives and needs of its users. In this sense, Self-adaptive Software (SaS) enables dealing with uncertainties through structural and/or behavioral changes at runtime. Based on the above scenario, our research group has developed some initiatives in this direction, which aim to support the SaS community and others interested in the development of SaS and Self-Apps (Self-adaptive Service-oriented Applications) supported by good practices of engineering software. Among of them, the design of a framework called DynaMS (Dynamic Deployment, QoS Metrics, and Semantic Search) resulted in some possibilities for future research. In this perspective, it is worth highlighting the improvement of the service monitoring system so that it can act more effectively, distinguishing services with a greater tendency to failures or degradation in QoS (Quality of Service). Therefore, the goal of this project is the development of a module, based on an intelligent approach that encompasses elements of self-awareness computing and reinforcement learning algorithms, to be coupled to the monitoring system of the DynaMS framework. Thus, it is intended, at the end of this project, to have a feasible solution for monitoring services in an optimized way, contributing to the balance and optimization of computational resources used in applications (Self-Apps).