International Conference on Simulations in Computing Nexus |
Foundation of Computer Science USA |
ICSCN - Number 2 |
May 2014 |
Authors: N. Mathiarasi, K. Abinaya, K. Kiruthikadevi |
5d81b851-e766-47f7-ab59-eda40d04306b |
N. Mathiarasi, K. Abinaya, K. Kiruthikadevi . Pipelining and Replica Optimization of Workflows for Infrastructure as Utility Grid. International Conference on Simulations in Computing Nexus. ICSCN, 2 (May 2014), 7-13.
Utility Grid is the collection of computer resources in a heterogeneous distributed environment which encourages researchers to investigate the benefits and drawbacks on executing workflows. While processing a large amount of workflows in utility grid environment, one of the most challenging problems is scheduling of workflows without replications/repetitions. Already, a two-phase approach called partial critical path approach to schedule the workflows in grid environment which aims to minimize the cost of workflow execution under user defined deadline. However when someone applies pipelining mechanism with such algorithm, the results obtained is not more optimized. In this paper, PCP algorithm is adopted with Replication Optimization phase (R-PCP) along with streaming pipelining mechanism. Applications such as scientific simulations, sensor network analysis generate huge amounts of data, which must be streamed efficiently. The streaming services must meet an application's quality of service (QoS) and this mechanism should guarantee that no data are lost during processing. The proposed mechanism operates on polynomial time complexity, which is suitable for optimizing large number of workflows. The main objective of this paper is to: implement pipeline mechanism in a large grid environment and making grid environment to work in an optimized way. The simulation results are shown on the paper gives promising results.