Cloud-based applications can achieve high performance and scalability provided through the infrastructure of cloud platforms. Distributed servers and technologies such as serverless computing further allow for distribution in terms of the network and for parallel execution. Distribution, however, does not imply decentralization, which additionally requires non-centralized coordination without trusted third parties. For example, independent network nodes may verify the execution using blockchain consensus. Despite advances in consensus algorithms, the execution of decentralized applications is challenging due to scalability and, when combined with cloud computing, the issue of authenticating distributed parties and execution states. This paper extends a model-based approach for execution and instance tracking on cloud platforms with scalability and authentication through certificates and blockchains. Thereby, distributed parties monitor and track execution states using models. The approach is extended by (a) an architecture for the authenticated and scalable distribution, execution, and tracking with executable models, and (b) a prototype evaluating feasibility for Amazon Web Services (AWS) and the Ethereum blockchain. Results indicate the potential for scalability and decentralization through executable models in the cloud with certificates and a blockchain.
Faculté des sciences économiques et sociales et du management