Identifying Dynamic Replication Strategies for a High-Performance Data Grid
GRID '01 Proceedings of the Second International Workshop on Grid Computing
Data Replication Strategies in Grid Environments
ICA3PP '02 Proceedings of the Fifth International Conference on Algorithms and Architectures for Parallel Processing
Enabling the Co-Allocation of Grid Data Transfers
GRID '03 Proceedings of the 4th International Workshop on Grid Computing
XORs in the air: practical wireless network coding
Proceedings of the 2006 conference on Applications, technologies, architectures, and protocols for computer communications
Complete and fragmented replica selection and retrieval in Data Grids
Future Generation Computer Systems
A List-Based Strategy for Optimal Replica Placement in Data Grid Systems
ICPP '08 Proceedings of the 2008 37th International Conference on Parallel Processing
Co-allocation in data grids: a global, multi-user perspective
GPC'08 Proceedings of the 3rd international conference on Advances in grid and pervasive computing
Replica selection on co-allocation data grids
ISPA'04 Proceedings of the Second international conference on Parallel and Distributed Processing and Applications
IEEE Transactions on Information Theory
A Random Linear Network Coding Approach to Multicast
IEEE Transactions on Information Theory
Hi-index | 0.00 |
We propose a new scheme for creating replica based on network coding in grid environments. For creating replica with network coding, there are two models: encoding between blocks of a replica and encoding between blocks of different replicas. The former is based on a butterfly-shaped topology, and the latter, in theory, based on any topology with a node more than one degree. Based on the latter model, furthermore, in practice, we propose a double-funnel-shaped topology. To evaluate our models, we extend NS2 by improving functions, modifying and adding classes. Compare to traditional strategies of creating replica, our experiment showed the strategy with network coding can improve data access time and I/O bottlenecks, low bandwidth consumption, or improve load balancing, increase robustness for being in support of departure of nodes.