Hybrid cloud computing QoS glitches

Abstract

The Hybrid Cloud Computing model has been growing extensively due to its Infrastructure as a Service (IaaS) architecture, customisation and cost benefits. The hybrid cloud services are measured based on the Quality of Service parameters defined by the public cloud vendors. These parameters (i.e. availability, scalability, latency etc.) vary from vendor-to-vendor, developing complexity and confusion on the grounds of methods of service assessments. A Cloud Service Level Agreement (SLA) lists the QoS provisions to be provided to the tenant, the objectives, and exclusions. Regardless of vendors promised uptimes and service metrics, the tenants are susceptible to the following threats: data governance, Denial of Services, multi-tenancy, etc. Cloud computing has often been compared as a utility, but the basic different between a utility and the cloud is the amount of risk involved with data protection, provisioning and control. Few cloud standards have been developed for standardizing the hybrid cloud model but since each public cloud vendor provides different applications and services, these standards do not resolve the existing cloud QoS issue. Since each enterprise implementing the cloud and vendor supplying the services is diverse, a customized Trio (Cloud-IT-Business) QoS model is required to resolve the business need. The authors have designed a model to resolve this existing cloud QoS issue, the abstraction of the model is detailed in this paper.