Comparação de algoritmos para o problema de escalonamento de tarefas em grades computacionais

Abstract

Task scheduling in computational grids is widely discussed in the scientific community. Since the tasks processed by a computer may have different or similar characteristics of complexity, and the machines themselves may have different or close computing resources, it is then necessary to schedule tasks so that all tasks can be executed. By the available machines in the best possible way, according to the selected metrics. In this sense, this work presents some algorithms already known in the scientific community. We compare this algorithms using different techniques, metrics, and scenarios to understand the advantages and disadvantages of them in each of the described scenarios.