Allocating optimized resources in the cloud by a viable SLA xodel

© 2016 IEEE. A cloud business environment comprises service providers and service consumers. Services are supplied through a Service Level Agreement (SLA) which defines all deliverables, commitments, obligations, QoS, violation penalties etc. that help a service provider and a service consumer to execute their business transactions. The primary aim of a service provider is to fulfill its commitment to a consumer by forming a viable SLA that wisely assigns the appropriate amount of resources to a requesting consumer. In this paper, we propose a viable SLA model that helps a service provider form a viable agreement with a consumer, based on its previous resource usage profile. The model uses a Fuzzy Inference System and takes the reliability and the contract duration of a consumer as input to calculate the suitability of this consumer, which is also used as input along with the risk propensity of a service provider to determine the amount of resources to offer to a consumer. We evaluate our approach and find that by using an optimized viable SLA model, providers are able to allocate an appropriate amount of resources to avoid an SLA violation