Functional Suitability and Evaluation Metrics for Autonomous Data Center Management: A Case Study of DCAMP

In response to the escalating demand for data usage and the imperative for stringent environmental sustainability, future data centers are being conceptualized for unconventional locations such as underwater, outer space, and remote areas. These boundaryless data centers necessitate sophisticated management solutions capable of autonomous operation with minimal human intervention. The novel Data Center Autonomous Management Platform (DCAMP) framework, leverages artificial intelligence, machine learning, automation, and contemporary software methodologies to construct an autonomous management platform with a focus on continuous improvement through robust knowledge management and human-in-the-loop based governance. This paper presents a comprehensive architectural evaluation of the DCAMP framework, assessing its quality attributes in accordance with ISO/IEC standards. Key performance indicators and metrics for managing boundaryless data centers are defined, and a usecase simulation evaluating the Functional Suitability characteristic of the DCAMP framework is discussed. Our findings demonstrate that DCAMP provides a compelling solution, offering high output accuracy and enhanced task efficiency, even amidst evolving requirements. The target audience for this paper includes researchers and practitioners in software and systems engineering. It is particularly relevant for data center managers, system administrators, infrastructure managers, SRE cloud engineers, and infrastructure researchers, offering them metrics and evaluation criteria to assess the effectiveness of autonomous data centers which need continuous adaptation.