Design and Analysis of Cloud-Based Architectures with KLAIM and Maude

摘要

Cloud computing is a modern paradigm for offering and utilizing distributed infrastructure resources in a dynamic way. Cloud-based systems are safety- and security-critical; they need to satisfy time-critical performance-based quality of service properties and to dynamically adapt to changes in the potentially hostile and uncertain environment they operate in. In this paper we propose the coordination language KLAIM and a composite actor approach for modelling Cloud-based architectures whereas for formally analyzing such architectures we use a rewriting-based approach. We specify the operational semantics of KLAIM in Maude, show how to realize KLAIM programs in a distributed implementation of Maude, and simulate and analyze three simple Cloud architectures with Maude and the Maude LTL model checker. Moreover, we report shortly on the Maude specification and analysis of three larger Cloud case studies using the composite actor model, where statistical model checking with the Maude-based tool PVeStA is successfully used for detecting bugs and performance issues and for analyzing a defense mechanism against distributed denial-of-service attacks.