Analysis of dependability aspects in data center systems integrating communication, power and cooling infrastructures

Abstract

Data centers are constantly growing, driving demands for new technologies such as cloud computing, e-commerce, which forced the availability of these systems 24 hours a day, 7 days a week under penalty of major losses. Our study aims to improve the dependability aspects of data center systems integrating the communication, power and cooling infrastructures considering different architectures. These architectures were defined based on the increment of redundancy regarding the power and cooling infrastructures considering a proposed communication infrastructure. In turn, this increase, which is based on the index of availability importance applied to the components of these infrastructures, considers adaptations of redundancy mechanisms widely used in system designs. Dependability models were created using the Reliability Block Diagram and stochastic Petri nets modeling mechanisms based on a hierarchical approach that considers the advantages and possibilities of both types of models. In addition, a study was carried out to analyze the availability of the different proposed architectures through the application of the hierarchical models.