Experiences on grid shared data programming
International Journal of Grid and Utility Computing
| Hi-index | 0.00 |
Besides the complexity of grid application program-ming, achieving high performance has become a real challenge inlarge scale grid systems mainly because of the high latencies andunpredictable communication paths. We follow the idea of shareddata programming in grid environments and we design a replica-tion component for a shared data programming framework withdynamic replication capabilities. We define several replicationpolicies that are dynamically used during the application life-cycle in order to minimize the communication cost and providebetter performance. We evaluate several application scenarios inorder to prove that dynamic policy based replication can improvethe overall application performance and reduce resource accesstime.