On the Co-existence of Service and Opportunistic Grids

Most computational grids currently in production are either service grids or opportunistic grids. While a service grid provides high levels of quality of service, an opportunistic grid provides computing power only on a best-effort basis. Nevertheless, since opportunistic grids do not require resources to be fully dedicated to the grid, they tend to assemble larger numbers of resources. Moreover, these grids cater very well for the execution of the so-called embarrassingly parallel applications, a type of application that is frequently found in practice. In this paper we present an approach for supporting the co-existence of a service grid and an opportunistic grid on the same infrastructure. The advantages of this hybrid infrastructure are twofold: firstly, the co-existence allows idle resources belonging to the service grid to be used in an opportunistic way; secondly, the provision of an opportunistic grid allows shared resources to be added to the infrastructure, a feature that turns out to be very important for consortia in which many of the member institutions cannot afford the provision of dedicated resources. The proposed approach is currently being implemented in the context of the EELA2 project.