Online Load Analysis For Automated Request-Quota Controlling In Clouds

The importance of efficient use of IT infrastructure capacities increases, as the growing amount of deployed IT systems increasingly requires resources. On the one hand, companies can outsource their IT demands to Cloud providers in order to only pay for the very amount of required computational power. On the other hand, cloud providers themselves need to be efficient as well, as it may be the deciding factor on their market positioning. This paper proposes a method to accept or refuse incoming requests based on a load estimation and load forecast. We will present the architecture and a general implementation of the so-called Quota-Control-Unit (QCU), followed by results of an extensive simulation. The evaluation will show that the introduced method is a promising approach of securing IT systems from overload while keeping the economic outcome at the possible maximum.