Towards the Requirement-Driven Generation and Evaluation of Hyperledger Fabric Network Designs

Abstract

Hyperledger Fabric (HLF) is an adaptable blockchain platform that enables the requirement-driven construction of cross-organizational distributed ledger networks. While this flexibility offers many advantages, it also introduces challenges in ensuring Fabric networks’ fit-for-purpose nature and extra-functional requirement adherence, e.g., with respect to fault tolerance. This challenge calls for the use of mathematically precise formal analysis techniques, tooling and models. In this paper, we propose a combination of model-driven engineering principles and diverse graph generation to validate Fabric network designs and to facilitate their design for dependability. Specifically, we present a Fabric network meta-model and a set of well-formedness requirements in the Graph-solver-as-aService tool Refinery. Uniquely, Refinery can check conformance on partial models, enabling analysis already during early design.