A temperature-aware resource management algorithm for holistic energy minimization in data centers

Reducing energy consumption is one of the key considerations for Cloud Providers (CPs). A traditional approach to address this issue is to formulate it into an optimization problem with QoS and server temperature constraints. In this paper, we develop a Temperature-Aware Resource Management (TARM) algorithm using Lyapunov Optimization theory. One advantage of our approach is that we use a “soft” (average) server temperature constraint instead of a “hard” (instant) one without impairing system reliability. We use a real world data center workload trace to evaluate our proposed algorithm, and simulation results show that our approach can at least save 6% of the overall energy consumption.