OutExaRD: Enhancing Resource Discovery for Support Dynamic and Interactive Events on Out-of-Distributed Exascale Systems

This study investigates the influence of dynamic and interactive events on the functionality of resource discovery within distributed Exascale systems. These events significantly impact resource discovery both before and after their occurrence. However, effectively analyzing these impacts is challenging due to the lack of a direct connection to the activator of resource discovery functionality. To address this challenge, we propose an innovative mechanism that establishes a connection structure defined by a wave vector, linking resource discovery functionality to its activator. This wave vector model is crucial in shaping resource discovery operations. It implements a searching system designed to create spatial dimensions, allowing resource discovery functionality to evaluate the impact of events on spatial generation within each searching system. This methodology facilitates a more effective analysis of events’ roles in managing their effects while minimizing disruption to system elements. Our research results demonstrate the efficacy of our approach, revealing that the wave vector model positively influenced resource discovery functionality in an impressive 93% of dynamic and interactive events, thereby optimizing the execution of resource discovery activities. This high success rate provides reassurance about the reliability of our approach.