From Grid5000
Revision as of 23:00, 20 November 2014 by Llefevre (talk | contribs) (Energy Efficient Large Scale Experimental Distributed Systems (Green))

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

Energy Efficient Large Scale Experimental Distributed Systems (Green)

Leaders: Laurent Lefevre (AVALON), Jean-Marc Menaud (ASCOLA)

As large scale distributed systems reach enormous size in terms of equipments, the energy consumption issue becomes one of the main challenges for large scale integra- tion. While energy aspects have been taken into account since a long time by researchers working on battery-based mobile and wireless distributed systems (ad-hoc networks, sen- sors. . . ), only few research activities are currently performed on fully connected equip- ments. However, these large-scale distributed systems contribute to energy increase and are more dependent on cooling systems. To face environmental crisis and reduce energy costs, the working group’s goal is to propose energy efficient realistic alternatives to sup- port high performance computing and networking usage in current and next generation large-scale distributed systems.

To address this challenge, different problems will be investigated:

  • Collecting energy usage of platforms at large scale level and exposing these information to users and administrators;
  • Benefiting of these logs in information systems and large scale distributed systems components (i.e. middleware, schedulers, resource managers) to optimize their usage;
  • Proposing energy aware software architecture able to autonomously react to energy related context changes (i.e. cooling information, management and administration policies, platform usage and configuration, applications and services deployment);
  • Investigating the design of energy oriented large scale autonomous decisions.

This working group focuses on energy aware software approaches able to reduce the energy consumption needed for high performance computing and networking operations in large scale distributed systems (datacenters, Grids and Clouds).

The Hemera working group Green has organized the GreenDays@Paris (May 31- June 1st, 2011) on the topic "Energy efficiency : how to apply our models and solutions in production infrastructures ?" (slides) and the GreenDays@Lyon (January 19-20, 2012) on the topic "Energy efficiency : how to monitor and impact on applications ?" (slides).