InterGrid: Internetworking Islands of Grids

Building a Virtualized Distributed Computing Infrastructure by Harnessing Grid and Cloud Technologies

InterGrid is inspired by the peering agreements established between Internet Service Providers (ISPs) in the Internet, through which ISPs agree to allow traffic into one another's networks. The architecture of InterGrid relies on InterGrid Gateways (IGGs) that mediate access to resources of participating Grids. The InterGrid also aims at tackling the heterogeneity of hardware and software within Grids. The use of virtualization technology can ease the deployment of applications spanning multiple Grids as it allows for resource control in a contained manner. In this way, resources allocated by one Grid to another are used to deploy virtual machines. Virtual machines also allow the use of resources from Cloud computing providers.

Abstract view of the software layers of the InterGrid

The InterGrid aims to provide a software system that allows the creation of execution environments for various applications (a) on top of the physical infrastructure provided by the participating Grids (c). The allocation of resources from multiple Grids to fulfil the requirements of the execution environments is enabled by peering arrangements established between gateways (b).

Application deployment

A Grid has pre-defined peering arrangements with other Grids, managed by IGGs and, through which they co-ordinate the use of resources of the InterGrid. An IGG is aware of the terms of the peering with other Grids; selects suitable Grids able to provide the required resources; and replies to requests from other IGGs. Request redirection policies determine which peering Grid is selected to process a request and a price at which the processing is performed. An IGG is also able to allocate resources from a Cloud provider. Figure 2 illustrates a scenario where an IGG allocates resources from an organisation's local cluster for deploying applications. Under peak demands, this IGG interacts with another that can allocate resources from a cloud computing provider.

Although applications can have resource management mechanisms of their own, we consider a case where the resources allocated by an application are used for the creation of a Distributed Virtual Environment (DVE), which is a network of virtual machines that runs isolated from other DVEs. Therefore, the allocation and management of the acquired resources is performed on behalf of the application by a component termed DVE Manager.

The Team Members


Stay Informed. Register Now!

Subscribe to Grid Computing Info Centre Newsletter
Browse Archive for past messages.
Cloud Computing and Distributed Systems (CLOUDS) Laboratory
School of Computing and Information Systems
The University of Melbourne, Australia