A comparison of receiver-initiated and sender-initiated adaptive load sharing
Performance Evaluation
Communication complexity for parallel divide-and-conquer
SFCS '91 Proceedings of the 32nd annual symposium on Foundations of computer science
Cilk: an efficient multithreaded runtime system
PPOPP '95 Proceedings of the fifth ACM SIGPLAN symposium on Principles and practice of parallel programming
A design study of the EARTH multiprocessor
PACT '95 Proceedings of the IFIP WG10.3 working conference on Parallel architectures and compilation techniques
Efficient load balancing for wide-area divide-and-conquer applications
PPoPP '01 Proceedings of the eighth ACM SIGPLAN symposium on Principles and practices of parallel programming
Satin: Efficient Parallel Divide-and-Conquer in Java
Euro-Par '00 Proceedings from the 6th International Euro-Par Conference on Parallel Processing
MANNA: Prototype of a Distributed Memory Architecture with Maximized Sustained Performance
PDP '96 Proceedings of the 4th Euromicro Workshop on Parallel and Distributed Processing (PDP '96)
The Natural Work-Stealing Algorithm is Stable
SIAM Journal on Computing
Scheduling multithreaded computations by work stealing
SFCS '94 Proceedings of the 35th Annual Symposium on Foundations of Computer Science
Heuristic static load-balancing algorithm applied to the fragment molecular orbital method
SC '12 Proceedings of the International Conference on High Performance Computing, Networking, Storage and Analysis
Hi-index | 0.00 |
Random stealing is a well-known dynamic load-balancing algorithm. However, for a large-scale cluster, the simple random stealing policy is no longer efficient because an idle node must randomly steal many times to obtain a task from another node. This will not only increase the idle time for all nodes but also produce a heavy network communication overhead. In this paper, we propose a novel dynamic load-balancing algorithm, Transitive Random Stealing (TRS), which can make any idle node obtain a task from another node with much fewer stealing times in a large-scale cluster. A probabilistic model is constructed to analyze the performance of TRS, random stealing and Shis, one of load balance policies in the EARTH system. Finally, by the random baseline technique, an experiment designed to compare TRS with Shis and random stealing for five different load distributions in the Tsinghua EastSun cluster convinces us that TRS is a highly efficient dynamic load-balancing algorithm in a large-scale cluster.