Toward a Next-Generation Data Center Management System: Enabling Application-Aware Integration

Abstract

Data centers are emerging as critical infrastructure in the digital era, providing the essential computing power needed to run smart applications within the modern communities. To meet the demands of these applications, data center must be designed as high-performance facilities, achieving both high energy efficiency and operational resilience. However, the growing volume of data collection and the need for real-time decisionmaking by applications running in the data center information and communication (ICT) devices, require an evolution of Data Center Management System (DCMS). The DCMS must be capable of optimizing the resilient operation of internal subsystems while seamlessly integrating external services. This paper introduces a conceptual framework for a next-generation DCMS, detailing its architectural layers, the inter-dependencies among subsystems, and integration with the application layer. The proposed framework is analyzed in the context of advanced artificial intelligence (AI) techniques, aiming to achieve key DCMS objectives such as minimizing energy consumption and carbon emissions, enhancing operational resilience, and optimizing resource allocation. Key challenges related to this integration are addressed, alongside an exploration of emerging trends and future research directions. This study provides a foundational framework for DCMS, highlighting the central role of data centers in enabling adaptive, efficient, and robust operation of the smart critical applications of today's digital society.