Smart Datacenter Electrical Load Model for Renewable Sources Management

Stéphane Caux (1) , Gustavo Rostirolla (2, 3) , Patricia Stolf (3)

1 LAPLACE-CODIASE - COmmande et DIAgnostic des Systèmes Electriques
2 LAPLACE - LAboratoire PLasma et Conversion d'Energie
3 IRIT-SEPIA - Système d’exploitation, systèmes répartis, de l’intergiciel à l’architecture

Nowadays, datacenters are one of the most consuming devices due to the increase in cloud, web-services and high performance computing demands all over the world. To be clean and without connection to the grid, datacenters projects tempt to feed electricity with renewable energy sources and storage elements. This power production needs an energy management providing power envelopes as a constraint to the datacenter management system. This paper presents an optimization module that optimizes the IT load under renewable energy constraints and outputs the power consumed by the computing resources of a datacenter. We are able to obtain a reduction of up to 73% in the tasks violations while respecting a given power envelope

Mots clés: Renewable energy; Renewable energy; Cloud computing; Workload; Scheduling; Energy Consumption

Domaines: Energie électrique

Stéphane Caux, Gustavo Rostirolla, Patricia Stolf. Smart Datacenter Electrical Load Model for Renewable Sources Management. International Conference on Renewable Energies and Power Quality (ICREPQ 2018), European Association for the Development of Renewable Energy, Environment and Power Quality (EA4EPQ), Mar 2018, Salamanca, Spain. pp.127-132, ⟨10.24084/repqj16.231⟩. ⟨hal-03886547v2⟩