A Methodology for Advanced Maintenance Management for Data Centers

The Internet of Things has an important impact on the Data Center (DC) market. The Gartner Group report 2013 estimates that the Internet of Things will include approximately 26 billion units, which shall be installed by 2020.... [ view full abstract ]

The Internet of Things has an important impact on the Data Center (DC) market. The Gartner Group report 2013 estimates that the Internet of Things will include approximately 26 billion units, which shall be installed by 2020. The emerging European market is one of the key drivers for this growth. Numerous new DCs will be built or will be refurbished to serve the growing need for digital infrastructure.

From a Facility Management (FM) perspective, the most important criteria to benchmark the efficiency of DC-operation are reliability (i.e. downtime) and cost. The limited downtime of a DC is based on the design as well as on the Operational and Maintenance Model. Depending on the classification and standard of the system from non-redundant up to fully fault tolerant, the downtime has a margin between 28.8 hours per year (99.671 % Availability) up to 0.4 hours per year (99.995 % Availability). This means that the combination of Reliability, Availability and Maintainability (RAM) is the most important key factor for an efficient DC-operation. Primarily, the human factor is most important to ensure a safe DC operation. The mentioned aspects drive many enterprises to consider preventive maintenance (PM) as an effective approach to enhance the reliability and the availability of DC. However, there is an increasing demand for harmonized and unified PM approach that enables maintenance labor to accomplish various maintenance activities effectively offers transnational procedures, which can be tuned according to the critical environmental conditions.

This paper demonstrates the critical role of maintenance in DC from operational and economic perspectives. Additionally, the PM approach is introduced. Research results based on different client practices are presented. Furthermore, the methodology for advanced Maintenance Management for DCs is described, providing the basis for benchmarking of costs and availability in DC.

Keywords:    Data Center, Maintenance, Facility Management, Reliability, Availability, Maintainability