Online Capacity Scaling Augmented With Unreliable Machine Learning Predictions

TL;DR

This paper introduces ABCS, an online capacity scaling algorithm for data centers that leverages machine learning predictions to optimize power and performance, maintaining robustness against prediction inaccuracies.

Contribution

The paper presents a novel adaptive algorithm that effectively uses black-box ML predictions for capacity scaling, ensuring near-optimal performance with theoretical guarantees.

Findings

ABCS is $(1+\varepsilon)$-competitive with accurate predictions.

ABCS maintains a bounded competitive ratio regardless of prediction accuracy.

Numerical experiments validate the theoretical performance of ABCS on real data.

Abstract

Modern data centers suffer from immense power consumption. As a result, data center operators have heavily invested in capacity scaling solutions, which dynamically deactivate servers if the demand is low and activate them again when the workload increases. We analyze a continuous-time model for capacity scaling, where the goal is to minimize the weighted sum of flow-time, switching cost, and power consumption in an online fashion. We propose a novel algorithm, called Adaptive Balanced Capacity Scaling (ABCS), that has access to black-box machine learning predictions. ABCS aims to adapt to the predictions and is also robust against unpredictable surges in the workload. In particular, we prove that ABCS is $(1+\varepsilon)$-competitive if the predictions are accurate, and yet, it has a uniformly bounded competitive ratio even if the predictions are completely inaccurate. Finally, we investigate the performance of this algorithm on a real-world dataset and carry out extensive numerical experiments, which positively support the theoretical results.