Study of execution model of permissioned blockchain using trusted serverless computing

Abstract

To overcome the performance limits inherent in blockchain systems, off-chaining models for compute and storage have been studied recently. Serverless computing offers increased elasticity and scalability in executing microservices in event-driven manner, making it proper fit for off-chain computation tasks. In this work, we present an off-chain execution model for permissioned blockchains leveraging trusted serverless computing. We propose a system model that extends execution engine of permissioned blockchains to effectively offload smart contract microservices to serverless functions. In addition, we show that hybrid execution model is proper for practical permissioned blockchain systems and propose a hybrid execution engine that provides optimized execution environment considering multiple factors. Finally, we identify the security threats inherent in serverless-based off-chain computation and demonstrate that incorporating trusted execution environment such as Intel SGX can help protect the security-sensitive data and code of smart contracts. The feasibility of our system is shown with experiments based on our case study about verification of supply chain transactions.