Thermal Awareness to Enhance Data Center Energy Efficiency

Data centers aim at provisioning on-demand processing, storage and networking capabilities in a reliable and scalable way. In this context, proper maintenance of IT equipment within DC premises is crucial as it ensures prolonged lifetime of servers and uninterrupted availability of resources. DC management teams’ sustainable operation effort comprises various approaches to directly and indirectly reduce DC energy consumption. Thermal management aims to reduce excess energy consumption by air cooling and compute systems. This paper focuses on the analysis of the exact temperatures in a real DC cluster rather than considering device setpoints or guidelines. An extensive statistical analysis of available thermal data collected by server-level sensors, global and local thermal metrics evaluation is conducted. It enables isolating possible risks engendered by potential negative covert cooling-related factors. The ultimate outcome of this research is to bring about improvement of DC thermal management for sustainable operations.