Towards an SLA-Based Service Allocation in Multi-cloud Environments

Cloud computing popularity is growing rapidly and consequently the number of companies offering their services in the form of Software-as-a-Service (SaaS) or Infrastructure-as-a-Service (IaaS) is increasing. The diversity and usage benefits of the IaaS offers are encouraging SaaS providers to lease resources from the Cloud instead of operating their own data centers. This helps them to get rid of the maintenance overheads and better satisfy their customers which are more demanding in terms of service requirements nowadays. Such evolutionary tendency is leading to the emergence of new ways of service provisioning in which relying on infrastructure services of a single Cloud provider is not sufficient. Namely, the need of using Cloud services from multiple Clouds with various quality attributes and pricing models has been raised recently. Although service allocation based on Service Level Agreement (SLA) has been well investigated in Cloud computing so far, the new upcoming issues regarding to utilize multiple Clouds has led to new challenges. This paper looks at the service selection and allocation in a Multi-Cloud, as a delivery model of multiple Clouds, from the perspective of SaaS provider. The proposed framework assists SaaS providers to find suitable infrastructure services which best satisfy their requirements while handling SLA issues. We present an overview of the complete system and discus how the services are selected and the corresponding SLAs are monitored to detect the SLA violations.