Section: New Results

Energy Efficiency of Large Scale Distributed Systems

Participants : Laurent Lefevre, Daniel Balouek-Thomert, Eddy Caron, Radu Carpa, Marcos Dias de Assunção, Jean-Patrick Gelas, Olivier Glück, Jean-Christophe Mignot, Violaine Villebonnet.

Energy Efficient Core Networks with SDN

This work [14], [15] seeks to improve the energy efficiency of backbone networks by providing an intra-domain Software Defined Network (SDN) approach to selectively and dynamically turn off and on a subset of links. We proposed the STREETE framework (SegmenT Routing based Energy Efficient Traffic Engineering) that represents an online method to switch some links off/on dynamically according to the network load. We have implemented a working prototype in the OMNET++ simulator and design a validation platform [15] based on NetFPGA and Raspberry equipment with SDN frameworks (ONOS).

Energy Proportionality in HPC Systems

Energy savings are among the most important topics concerning Cloud and HPC infrastructures nowadays. Servers consume a large amount of energy, even when their computing power is not fully utilized. These static costs represent quite a concern, mostly because many datacenter managers are over-provisioning their infrastructures compared to the actual needs. This results in a high part of wasted power consumption. In this work [25], [24], [23] , we proposed the BML (“Big, Medium, Little”) infrastructure, composed of heterogeneous architectures, and a scheduling framework dealing with energy proportionality. We introduce heterogeneous power processors inside datacenters as a way to reduce energy consumption when processing variable workloads. Our framework brings an intelligent utilization of the infrastructure by dynamically executing applications on the architecture that suits their needs, while minimizing energy consumption. Our first validation process focuses on distributed stateless web servers scenario and we analyze the energy savings achieved through energy proportionality. This research activity is performed with the collaboration of Sepia Team (IRIT, Toulouse) through the co-advising of Violaine Villebonnet.

Energy-Aware Server Provisioning

Several approaches to reduce the power consumption of datacenters have been described in the literature, most of which aim to improve energy efficiency by trading off performance for reducing power consumption. However, these approaches do not always provide means for administrators and users to specify how they want to explore such trade-offs. This work [11] provides techniques for assigning jobs to distributed resources, exploring energy efficient resource provisioning. We use middleware-level mechanisms to adapt resource allocation according to energy-related events and user-defined rules. A proposed framework enables developers, users and system administrators to specify and explore energy efficiency and performance trade-offs without detailed knowledge of the underlying hardware platform. Evaluation of the proposed solution under three scheduling policies shows gains of 25% in energy-efficiency with minimal impact on the overall application performance. We also evaluate reactivity in the adaptive resource provisioning This research activity is performed with the collaboration of NewGen SR society through the co-advising of Daniel Balouek-Thomert.

Improving Energy Re-Usage of Large Scale Computing Systems

The heat induced by computing resources is generally a waste of energy in supercomputers. This is especially true in very large scale supercomputers, where the produced heat has to be compensated with expensive and energy consuming cooling systems. Energy is a critical point for future supercomputing trends that currently try to achieve exascale, without having its energy consumption reaching an important fraction of a nuclear power plant. Thus, new ways of generating or recovering energy have to be explored. Energy harvesting consists in recovering wasted energy. ThermoElectric Generators (TEGs) aim to recover energy by converting wasted dissipated energy into usable electricity. By combining computing units (CU) and TEGs at very large scale, we spotted a potential way to recover energy from wasted heat generated by computations on supercomputers. In this work [30], [20], we study the potential gains in combining TEGs with computational units at petascale and exascale. We explored the technology behind TEGs, and finally our results concerning binding TEGs and computational units in a petascale and exascale system. With the available technology, we demonstrate that the use of TEGs in a supercomputer environment could be realistic and quickly profitable, and hence have a positive environmental impact.