Section: New Results
Optimized protocols implementation and networking equipments
Locating Virtual Infrastructures: Users and InP Perspectives
Participants : Paulo Gonçalves, Guilherme koslovski.
This is a joint work with Pascale Vicat-Blanc (Lyatiss) and Sébastien Soudan (Lyatiss).
The Cloud Computing wave consolidates the on-demand provisioning of configurable virtual machines. Recent projects have proposed the extension of the original IaaS paradigm to provide dynamic virtual networks to interconnect virtual IT resources, composing Virtual Infrastructures (VIs). In this new scenario, users with different objectives and expectations can rent dynamically provisioned virtual infrastructures to execute their applications during a given time slot. VIs can be allocated anywhere on top of a distributed and virtualized substrate. This decoupling from the geographical location introduces concerns such as a latency increase in network communications (user’s perspective), and the fragmentation of physical resources (Infrastructure Provider’s - InP - perspective). This context motivates efforts to investigate and deploy new models and tools which consider the geographical location of virtual infrastructures. Our work concentrates on the allocation of VIs guided by both the user’s and the InP’s constraints. We propose a formulation of the allocation problem considering the user’s expectations as well as the physical-substrate provider’s goals. Our initial experiments demonstrate that it is possible to improve the quality of the virtual-infrastructure allocation (user perspective) while simultaneously decreasing the physical substrate’s fragmentation and the substrate’s cost.
Energy-efficient reservation infrastructure for large-scale distributed systems
Participants : Anne-Cécile Orgerie, Laurent Lefèvre, Guérin-Lassous Isabelle.
Over the past few years, the energy consumption of Information and Communication Technologies (ICT) has become a major issue. Nowadays, ICT accounts for 2% of the global CO2 emissions, an amount similar to that produced by the aviation industry. Large-scale distributed systems (e.g. Grids, Clouds and high-performance networks) are often heavy electricity consumers because – for high-availability requirements – their resources are always powered on even when they are not in use. Reservation-based systems guarantee quality of service, allow for respect of user constraints and enable fine-grained resource management. For these reasons, in the context of Anne-Cecile Orgerie Phd (defended in September 2011), we proposed an energy-efficient reservation framework to reduce the electric consumption of distributed systems and dedicated networks. The framework, called ERIDIS, is adapted to three different systems: data centers and grids, cloud environments and dedicated wired networks. By validating each derived infrastructure, we show that significant amounts of energy can be saved using ERIDIS in current and future large-scale distributed systems [54] .
Energy efficiency in exascale infrastructures
Participants : Mehdi Diouri, Olivier Gluck, Laurent Lefevre.
Joint work with F. Cappello (JLPC, joint laboratory between INRIA and NCSA).
In Diouri's PhD, we address the issue of energy efficiency for exascale supercomputers. We first proposed a green architecture for exascale systems gathering some new solutions to “consume less” energy and to “consume better”. This architecture involves interactions with the different actors interfering directly or indirectly with the supercomputer: its user, its administrator, its resource manager and the energy supplier. Then we were interested into leaning on this green architecture in order to propose some green services that will be offered for applications that will run on exascale systems. Our approach consists in evaluating the power overhead induced by some existing services, and by proposing a green version of these services that takes into account the constraints imposed by the different actors involved. In 2011, we specifically aimed to apply our approach for fault tolerance protocols in their normal functioning stage and in case of failure. [38]
Energy profiling and green leverages for high performance computing applications
Participants : Ghislain Landry Tsafack, Jean-Patrick Gelas, Laurent Lefevre.
Ghislain Landry TSAFACK CHETSA has started his PhD in january 2011, within the framework of INRIA HEMERA project, on: “Energy profiling and green leverages for high performance computing applications” (co-advisement with Jean-Marc PIERSON and Patritia STOLF from IRIT). During the course of this first year, we have investigated the possibility of characterizing distributed applications considering their energy / power profile. We first carried out a set of experiments for a better understanding of the application's behavior and the impact that this behavior may have on its power consumption. Results led us to the assumption that any individual application run can be represented as a sequence of basic operations including computation, memory accesses, disk and network accesses over a given time period. We next relay on that assumption to define application's energy profile. Application's energy profile helps to prevent the fallout of any action that may be taken to reduce its power usage. To guarantee reasonable results, i.e., reduce energy with less performance degradation, we designed an energy prediction model capable of predicting power usage of a wide range of high performance computing (HPC) applications.
Towards virtualized home gateways
Participants : Jean-Patrick Gelas, Laurent Lefevre, Anne-Cecile Orgerie.
Joint work with Dino Lopez Pacheco (Univeristy of Nice) and Referi Assefa (Addis Abeba University, Ethiopia).
Virtualizing services located on end to end parts of the networks and making them available for a large number of applications and users is now becoming a real challenge. Within the scope of the GreenTouch project, we are exploring models, simulations tools (ECOFEN) and software prototypes able to demonstrate the impact of such approach in terms of energy reduction.