FEC-ARRC: Optimization of the Fraunhofer Edge Cloud with automatic recommendations for resource configuration
ARRC optimizes the Fraunhofer Edge Cloud through seamless OpenStack integration, explainable AI, and automated rightsizing via GitOps. In the FEC-ARRC project, we are integrating the Automatic Recommender for Resource Configuration, or ARRC for short, into the OpenStack ecosystem of the Fraunhofer Edge Cloud (FEC). The FEC is a productively operated reference system that shows companies how edge and cloud platforms work in everyday life. ARRC uses explainable artificial intelligence, time series analysis, and multi-agent game theory methods. This enables the system to continuously and decentralizedly evaluate the utilization and configuration of the platform, make rule-compliant recommendations on the correct size of resources, and automatically implement approved changes using the GitOps operating model.