Inderscience Publishers

A case study of transport protocols to improve the execution of applications in virtual organisations utilising multicluster network configurations

0

Distributed team-based collaboration is an established trend in modern organisations. Commodity-off-the-shelf cluster architectures have become common in providing computational resources for a large number of complex applications. Clusters are often gathered into multicluster network configurations within and among organisations. These configurations can be considered as grid environments, where computational resources are scattered over a wide area network. These grid environments face a number of well-known challenges, including the heterogeneity of computational resources (e.g., different utilisation policies, diverse operating systems and processor architectures), different programming paradigms and network protocols. Network protocols are especially important, because these have a significant impact on the execution of distributed global applications. However, network configurations in many organisations do not utilise protocols aimed at ensuring the performance of these global applications. In this paper we present a case study of the performance impact of three important transport protocols on the execution of applications within real Virtual Organisations (VOs) when utilising multicluster network configurations. We report on the performance of these three different protocols in the context of throughput, latency and number of flows.

Customer comments

No comments were found for A case study of transport protocols to improve the execution of applications in virtual organisations utilising multicluster network configurations. Be the first to comment!