Two-Phase Industrial Manufacturing Service Management for Energy Efficiency of Data Centers

Data-driven industrial manufacturing services are proliferating. They use large amounts of data generated by Industrial-Internet-of-Things devices for intelligent services to end-service-users. However, cloud data centers hosting these services consumes huge amount of energy and contributing to high operational cost. To address this issue, this paper proposes an energy-efficient resources allocation framework for cloud services. It operates in two phases. Firstly, a multi-threshold based host CPU utilization classification scheme is developed to classify hosts into four groups. It is designed through analyzing the CPU utilization data using the least median squares regression model. Thereby, the scheme limits search space, thus reducing time complexity. Secondly, with a metaheuristic search, an energy and thermal-aware resource allocation method is developed to find an energy-efficient host for allocating resources to services. From real data center workload traces, extensive experiments show that our frame-work outperforms existing baseline approaches with 6.9%, 33.75%, and 34.1% on average in terms of temperature, energy consumption, and service-level-agreement violation respectively.