ESAVE: Estimating Server and Virtual Machine Energy

TL;DR

ESAVE is a machine learning approach that accurately estimates server and virtual machine energy consumption using minimal hardware attributes, aiding sustainable software deployment without intrusive measurements.

Contribution

Introduces ESAVE, a novel ML-based method for non-intrusive energy estimation of servers and VMs with high accuracy using few hardware features.

Findings

Average prediction error of around 12%

Effective with minimal hardware attributes

Non-intrusive energy estimation method

Abstract

Sustainable software engineering has received a lot of attention in recent times, as we witness an ever-growing slice of energy use, for example, at data centers, as software systems utilize the underlying infrastructure. Characterizing servers for their energy use accurately without being intrusive, is therefore important to make sustainable software deployment choices. In this paper, we introduce ESAVE which is a machine learning-based approach that leverages a small set of hardware attributes to characterize a server or virtual machine's energy usage across different levels of utilization. This is based upon an extensive exploration of multiple ML approaches, with a focus on a minimal set of required attributes, while showcasing good accuracy. Early validations show that ESAVE has only around 12% average prediction error, despite being non-intrusive.