MagPIe: MPI's collective communication operations for clustered wide area systems
Proceedings of the seventh ACM SIGPLAN symposium on Principles and practice of parallel programming
Wide-area implementation of the message passing interface
Parallel Computing - Special issue on applications
A grid-enabled MPI: message passing in heterogeneous distributed computing systems
SC '98 Proceedings of the 1998 ACM/IEEE conference on Supercomputing
Efficient Collective Communication on Heterogeneous Networks of Workstations
ICPP '98 Proceedings of the 1998 International Conference on Parallel Processing
ECO: Efficient Collective Operations for Communication on Heterogeneous Networks
IPPS '96 Proceedings of the 10th International Parallel Processing Symposium
HCW '99 Proceedings of the Eighth Heterogeneous Computing Workshop
Bandwidth-Efficient Collective Communication for Clustered Wide Area Systems
IPDPS '00 Proceedings of the 14th International Symposium on Parallel and Distributed Processing
Exploiting Hierarchy in Parallel Computer Networks to Optimize Collective Operation Performance
IPDPS '00 Proceedings of the 14th International Symposium on Parallel and Distributed Processing
On the Impact of Communication Complexity on the Design of Parallel Numerical Algorithms
IEEE Transactions on Computers
A Peer-to-peer Architecture forWorkflow in Virtual Enterpris
QSIC '05 Proceedings of the Fifth International Conference on Quality Software
Topology-Based hypercube structures for global communication in heterogeneous networks
Euro-Par'05 Proceedings of the 11th international Euro-Par conference on Parallel Processing
| Hi-index | 0.00 |
Binomial trees have been used extensively for broadcasting in clusters of workstations. In the case of heterogeneous non-dedicated clusters and grid environments, the broadcasting occurs over a heterogeneous network, and the performance obtained by the broadcast algorithm will depend on the organization of the nodes onto the binomial tree. The organization of the nodes should take into account the network topology, i.e., the communication cost between each pair of nodes. Since the network traffic, and consequently the latency and bandwidth available between each pair of nodes, is constantly changing, it is important to update the binomial tree so that it always reflects the most current traffic condition of the network. This paper presents and compares strategies to dynamically adapt a binomial tree used for broadcasting to the ever-changing traffic condition of the network, including accommodating nodes joining the network at any time and nodes suddenly leaving.