PARALLEL DATA LAB 

PDL Abstract

A Call for Research on Storage Emissions

HotCarbon’24, July 9, 2024, Santa Cruz, CA.

Sara McAllister, Fiodar Kazhamiaka†, Daniel S. Berger†, Rodrigo Fonseca†, Kali Frost†, Aaron Ogus†, Maneesh Sah†, Ricardo Bianchini†, George Amvrosiadis, Nathan Beckmann, Gregory R. Ganger

Carnegie Mellon University
†Microsoft Azure

http://www.pdl.cmu.edu/

Major cloud providers have committed to lowering carbon emissions by 2030 across their datacenters, and research has contributed many ideas on how this may be achieved. However, a major contributor to datacenter emissions has not received enough attention: storage. Storage — everything from file storage to inter-application messaging in datacenters — causes 33% of operational emissions and 61% of embodied emissions in Azure’s general-purpose cloud, based on a recent study. This paper identifies key sources of both operational and embodied emissions within distributed storage in datacenters. We also discuss strategies to reduce storage emissions and their challenges due to storage’s fundamentally stateful nature.

FULL TR: pdf