Analysis of Environmental Data in Data Centers

Optimization of data center performance for reduction in power and cooling costs is crucial to the growth of IT services both in enterprise and consumer sector. Significant work has been conducted in the past to understand the thermodynamic variables and parameters that influence the environmental conditions of the rooms vis-a-vis the computer room air conditioning (CRAC) units, racks and servers. A considerable amount of data had been collected from environmental sensors located at various locations within data centers, measuring the supply and return air temperatures in the CRAC units and both inlet and outlet temperatures in the racks. This work describes the analysis of environmental data collected from such sensors to discover and classify trends and patterns and relationships within temperatures and air flow data in a data center. Data mining techniques have been used to extract knowledge from historical data to improve the dynamic monitoring and control of data center power and cooling resources. Initially exploratory data analysis (EDA) techniques were used for reduction of data, visualization of deterministic behavior and identification of normal or abnormal environmental behavior in the control process. Since the most important part of the analysis is to understand the influence of the CRAC units over the racks, Principal Components Analysis (PCA) was used to capture the variables that contain most of that information.