Towards predictive cost models for cloud ecosystems: Poster paper

The rapid increase in the demand for cloud computing is driving a more recent increasing move towards multi-cloud environments by infrastructure providers. However, one of the main challenges for infrastructure providers is in sustaining profitability while provisioning services and components in such multi-cloud environments. This poster paper presents work towards predictive cost modeling for cloud ecosystems. Both real-time and historical analysis of service, virtual and physical resource information gathered from cloud infrastructure is used to drive both economic assessment and prediction models which can anticipate future service economic trends and the effect on total cost of ownership at the infrastructure provider side. This aids driving decisions to be made based on increased or reduced costs, which aids decision-support to effectively fulfil SLA requirements.