Model of load balancing and scheduling in Cloud computing

While implementing Cloud based platforms it is complicated to determine the adequate quantity of resources in order to satisfy variable workload with peaks. Some projects are lost because of under provisioning of Cloud resources, as well as costly resources are wasted during nonpeak periods. It is known that not all projects must be processed on time, i.e. some of them may be postponed for a certain time. Within this context, it is presumed that postponement's opportunity can reduce the probability of projects not to be executed, at the same time to determine the lower quantity of own resources by taking into account minimal costs. Such strategy of workload balancing and projects scheduling is validated by the experiment.