Load Information Sharing Policies In Communication-Intensive Parallel Applications
One usage of Grid infrastructures is to perform parallel computing of scientific applications, most of the time related to hard sciences (physics, chemistry, biology). To exploit parallelism most of these applications are intensive communicated in data and synchronisation messages. On this context, grid systems have to take in account to not interfering with the normal execution of applications. Starting from this idea, in this article we present a study of information sharing policies used by load-balancing algorithms developed for the middleware ProActive, analyzing the performance scalability of: response time (time of reaction against instabilities) and bandwidth, from a communication-intensive application context. We divided the policies into: Centralized or Distributed oriented; and Eager or Lazy load information sharing. Our experimental results show that Eager Distributed oriented policies have better performance (response time and bandwidth usage).
KeywordsDynamic load balancing Communication-intensive parallel applications Load information sharing policies Load information collection
Unable to display preview. Download preview PDF.
- Petra Berenbrink, Tom Friedetzky, and Leslie Ann Goldberg. The natural work-stealing algorithm is stable. In IEEE Symposium on Foundations of Computer Science, pages 178-187, 2001.Google Scholar
- Javier Bustos-Jimenez, Denis Caromel, Alexandre Di Costanzo, Mario Leyton, and Jose Piquer. Balancing active objects on a peer to peer infrastructure. In Proceedings of XXV International Conference of SCCC, Valdivia, Chile. IEEE CS Press, November 2005.Google Scholar
- Steve J. Chapin, Dimitrios Katramatos, John Karpovich, and Andrew S. Grimshaw. The Legion resource management system. In Dror G. Feitelson and Larry Rudolph, editors, Job Scheduling Strategies for Parallel Processing, pages 162-178. Springer Verlag, 1999.Google Scholar
- Wilfired Klauser Denis Caromel and Julien Vayssiere. Towards seamless computing and metacomputing in java. Concurrency Practice and Experience, 1998.Google Scholar
- Elisa Heymann, Miquel A. Senar, Emilio Luque, and Miron Livny. Adaptive scheduling for master-worker applications on the computational grid. In GRID, pages 214-227, 2000.Google Scholar
- Fabrice Huet, Denis Caromel, and Henri Bal. A high performance java middleware with a real application. In Proc. of High Performance Computing, Networking and Storage (SC2004), Pittsburgh, USA, 2004.Google Scholar
- Miron Livny Michael Litzkow and Matt Mutka. Condor - a hunter of idle workstations. In Proc. of 8th International Conference on Distribuited Computing Systems, pages 104-111, 1998.Google Scholar
- J.L. Bosque Orero, D. Gil Marco, and L. Pastor. Dynamic load balancing in heteroge- neous clusters. In Proc. of IASTED International Conference on Parallel and Distributed Computing and Networks, 2004.Google Scholar
- Niranjan G. Shivaratri, Phillip Krueger, and Mukesh Singhal. Load distributing for locally distributed systems. IEEE Computer, 25(12):33-44, 1992.Google Scholar
- W. Zhu, C. Steketee, and B. Muilwijk. Load balancing and workstation autonomy on amoeba. Australian Computer Science Communications, 17(1):588-597, 1995.Google Scholar