Efficient Resource Allocation for Web Applications Hosted in the Cloud by means of Weighted Multi-objective Linear Programming

• Valter Rogério Messias USP
• Júlio Cezar Estrella USP
• Ricardo S. Ehlers USP

Resumo

In a cloud computing environment, companies can allocate and de-allocate computing resources according to demand. However, this task does not happen instantaneously. There is a delay, which may take minutes, between the request for a new resource and it be ready for use. To resolve this problem we need forecast the future demand for then allocate the required amount of resources. As there is no forecasting model which is appropriate in all cases, we propose in this paper, the combination of different forecasting models by means of weighted multi-objective linear programming. In this way, we intend to create a generic method, able to optimize the resource allocation for different web applications, with different demand types. The results obtained by means of simulation show that our proposal achieves this goal.