Abstract
Cloud computing is a new way of computing that gave users an opportunity to share resources in a virtual storage. Therefore, cloud systems provide universal environment with a considerable number of Virtual Machines (VMs) that is available to many users who can access them online. This form of access makes cloud systems weaker than physical networks, and consequently, they are easy goals for attackers. In order to reduce the number of attacks, any malicious behavior should be rapidly discovered and stopped. In this paper, there is a continuation of the research that focus on the discovery of malicious behavior symptoms. The main contribution of this paper is the simulation analysis of algorithm Reduce Max with respect to its delay in monitoring cloud systems with different demand functions according to the appearance of symptoms that should be on a random occurrence.