CarbonScaler: Leveraging Cloud Workload Elasticity for Optimizing Carbon-Efficiency

TL;DR

CarbonScaler is a novel cloud workload management approach that dynamically adjusts server allocation based on real-time carbon intensity, significantly reducing the carbon footprint of batch jobs in cloud environments.

Contribution

The paper introduces the concept of carbon scaling for cloud workloads, developing a greedy algorithm and implementing a prototype to optimize carbon emissions during job execution.

Findings

Achieves 51% carbon savings over carbon-agnostic methods

Outperforms suspend-resume policies by 37% in carbon reduction

Surpasses static scaling policies by 8% in carbon efficiency

Abstract

Cloud platforms are increasing their emphasis on sustainability and reducing their operational carbon footprint. A common approach for reducing carbon emissions is to exploit the temporal flexibility inherent to many cloud workloads by executing them in periods with the greenest energy and suspending them at other times. Since such suspend-resume approaches can incur long delays in job completion times, we present a new approach that exploits the elasticity of batch workloads in the cloud to optimize their carbon emissions. Our approach is based on the notion of "carbon scaling," similar to cloud autoscaling, where a job dynamically varies its server allocation based on fluctuations in the carbon cost of the grid's energy. We develop a greedy algorithm for minimizing a job's carbon emissions via carbon scaling that is based on the well-known problem of marginal resource allocation. We implement a CarbonScaler prototype in Kubernetes using its autoscaling capabilities and an analytic tool to guide the carbon-efficient deployment of batch applications in the cloud. We then evaluate CarbonScaler using real-world machine learning training and MPI jobs on a commercial cloud platform and show that it can yield i) 51% carbon savings over carbon-agnostic execution; ii) 37% over a state-of-the-art suspend-resume policy; and iii) 8% over the best static scaling policy.