CONICET | Buscador de Institutos y Recursos Humanos

Grid Computing proposes unlimited access to different computational resources in a transparent way. High performanceexecution in grid environments is virtually impossible without timely access to accurate and up-to-date informationrelated to distributed resources and services. Due to inherent difficulty of testing the different information propagationpolicies in real grid infrastructures, several simulation frameworks arose to help in this issue. In this work, we presentGrid Matrix, an extension to one of the most used grid simulation tools (SimGrid2) to focus on the propagation of moni-toring and resource information allowing the creation of virtual grid infrastructures. This extension enables GUI editingof network topology and provides the feature of scripting to define simulation details based on the newly developedC ++ and Python bindings of SimGrid2 API. As a case study, Grid Matrix was used to test four different policies: hier-archical, super-peer, best-neighbor and random. The simulated scenario consisted of 96 master nodes based on the realTeragrid infrastructure as was publicly available at the time of writing this paper. We introduce three metrics that cap-ture and summarize the information propagation behavior: LIR, GIR and GIV. LIR captures the local behavior quantifyingthe amount of up-to-date information in each node. GIR evaluates the global information state in the whole networkaveraging the LIR values, while GIV measures the variability of LIR. In the presented case, the best results in terms ofthe proposed metrics were attained by the hierarchical policy, followed by super-peer which outperformed random andbest-neighbor. The modern and modular design of the scripting features included in Grid Matrix, in close conjunctionwith the user friendly GUI happened to be a very powerful tool for the evaluation of new propagation policies ofresource information.