MRI Consortium: Development of Instrumentation for Measuring the Dependability and Quality of Cloud Computing Systems

点击翻译按钮获取中文摘要

This project, aiming to develop instrumentation that enables rigorous experimental quality and dependability assessment of cloud computing systems, including cloud service-oriented architecture(SOA) systems, dramatically enhances the validation capability of cloud platform services and cyberinfrastructure. Cloud platforms must meet specified service level agreements (SLA) and provide mechanisms to support the minimization of the power used in cloud data centers without adversely affecting the quality of service (QoS). Also, some safety-critical emerging cloud applications, such as health-care and transportation applications, must meet stringent dependability requirements, that include high availability, reliability, performance, resilience, safety, and security. The proposed work establishes a high performance, metrics-driven instrument enabling rigorous evaluation for the following research projects on cloud performance: - Assurance of QoS in Service Clouds,- Optimization of Dependability in Evolving Clouds,- Mobile Device Power Management in Service Clouds,- Achievement of Highly Secure Compositions in Service Clouds, and- Testing and Validation of Secure Hardware-Software Architectures.The work aims to develop a flexible instrument for validation and verification of cloud performance, seen to be a key element in the future of computing services. In recent years, computer services have been moving to remote virtual machines ?in the cloud.? The commercial success of this new business model is based on service quality guarantees, which are becoming increasingly hard to verify as the complexity and volume of cloud services increase. Testbeds of this type are critical to the economic and research vitality of the academic and industrial communities.Broader Impacts: This instrumentation substantially raises the quality, scale, and scope of experimental research in dependability enhancement methods for cloud computing systems. It helps foster a strong scientifically based experimental paradigm for computing and provides capabilities that will greatly enhance the quality of undergraduate and graduate courses in computer science and engineering, as well as senior design projects by enabling students to work on realistic platforms. The instrumentation enhances the facilities of an NSF I/UCRC and will be available to researchers at the participating institutions and industrial partners.