Cloud computing is a game-changing paradigm in modern computing, offering a better utilization of IT resources and at the same time reducing the operational costs and increasing the elasticity of the IT infrastructure. Clouds exist in various shapes and dimensions, from general purpose public clouds having virtually unlimited resources, to small private clouds used inside enterprises for specific business goals. These private clouds have limited resources in terms of processing power and IO capability and therefore, an effective scheduling of workloads becomes very important. Another characteristic of small, private clouds is that they are built for a specific purpose, allowing an accurate prediction of the workloads that are going to be scheduled. This paper presents the characteristics of a private cloud used for e-Learning purposes along with a genetic algorithm that optimizes the scheduling of the e-Learning workloads according to a set of conditions that are imposed by the underlying virtualization technology such as memory over-commitment and IOPS rate distribution.