Model for SLA-Based VM Scheduling in Federated Cloud Environments

The future will see more and more cloud service providers joining forces as cloud federations so as to offer their customers a wider range of services whilst optimizing usage of their resources. Service level agreements (SLAs) form the basis for these federations by defining the quality of services between the providers themselves and between the provider and the customer. The distributed, SLA-based scheduling of virtual machines (VMs) presents a major challenge. For this reason a general model is developed that will allow us to explore different VM scheduling strategies in distributed cloud federations. This model can be used as a template to investigate complex cloud federation scenarios using the multi-level scheduler presented here.