Inter-Clouds: Grid Prosthesis for Workflow Systems

Nowadays cloud infrastructures have garnered popularity due to facility of outsourcing resources. Infrastructure as a Service clouds provide a exible way to allocate remote distributed resources, but lack mechanisms to configure software dependencies and to execute and manage applications. To make most use of cloud infrastructures, scientific applications should be able to harness cloud resources from different providers to maximize its performance. Here we show how a Scientific Work ow Management Systems (SWMS) is coupled with an Inter-Cloud (IC) infrastructure and scales out grid resources onto cloud so as to scale up work ows. We present a work ow executing on top of grid and IC. We
show that scaling happens at two levels: the work ow layer, where the work
ow tasks are scaled independently and the resource layer where the resources are scaled on-demand to match the demand from the application.