Description:
The use of grid systems for distributing and managing resources such as computing power and data storage has become increasingly widespread in recent years. However, as the demand for these resources continues to grow, the capacity of traditional grid systems to meet this demand has become a concern. When dealing with the constantly expanding system scale and its many uncertainties, traditional model-based techniques are becoming unsuitable. A better alternative to these techniques involves considering data-driven control (DDC) methodologies. In this paper, we begin by reviewing the current state of the art in DDC usage in grid systems in monitoring, improving, error detection, etc. with a particular focus on improving host capacity. We then describe our proposed approach, which involves improving the host capacity of grid systems using historical data. Finally, we present experimental results demonstrating the effectiveness of our approach and discuss its potential impact and future directions.