The Research Of Metrics And Measurement Method On Quality Of Cloud Services

Cloud computing services provide open standards, scalable system and service oriented architecture, so the cloud services providers can provide reliable and on-demand services for users in a flexible and economic way. These unique advantages have been made more and more users to use cloud services to build their own application systems. The reliability, performances etc. of the application systems are inseparable with the quality of cloud services. In order to ensure the normal running of these systems, cloud services providers must provide a certain quality assurance. Well, how the users can know what kind of services they have gotten and whether the services have satisfied the negotiated quality? So we need to measure the quality of cloud services. In this paper, we proposed a method and distributed framework on the quality of cloud services measurement.This paper mainly studies the metrics and measurement method on quality of services (QoS) in the cloud computing environment, thus can made services providers to provide services which meet the requirements of users and attain reputation for services providers. Firstly, this paper analyzes the features of cloud computing, the resources of cloud services and the definition, development status, existing problems of the SLA parameters, then proposes a measurement method, which measures the cloud services from the four following aspects of cloud services:availability, performance, throughput rate and utilization rateThe resources of cloud services not only include resources of local services, but also include the resources of remote services, so the measurement framework on the quality of cloud services must adapt distributed technology. This paper applied SOA architecture and technologies to the measurement framework, so effectively met the scalable requirement. In the framework, all the services were registered to the services bus, all the measurement information of cloud services were collected by the plug-ins, the QoS data were computed with the computing model of QoS parameters, and then compared the QoS data with the quality negotiated in the SLA. When the quality was not satisfied, the relevant module would give an alarm; notify the providers to restore the negotiated quality. At the same time, the framework provided a service management platform. Users can conveniently query data of QoS, management the computing model of QoS parameters, dealing with alarms etc.This paper just made some useful attempts to the method and framework on the quality of cloud services measurement, so we need to do some further study for the service availability and measurement framework.