HotStuff is a Byzantine fault-tolerant state machine replication protocol that incurs linear communication costs to achieve consensus. This linear scalability promoted the protocol to be adopted as the consensus mechanism in permissioned blockchains. This paper discusses the architecture, testing, and evaluation of our extensible framework to implement HotStuff and its variants. The framework already contains three HotStuff variants and other interchangeable components for cryptographic operations and leader selection. Inspired by the Twins approach, we also provide a testing framework for validating protocol implementations by inducing Byzantine behaviors. Test generation is protocol-agnostic; new protocols can execute the test suite with little-to-no modifications. We report relevant insights on how we benefited from Twins for validation and test-driven development. Leveraging our deployment tool, we evaluated our implementation in various configurations.
Systems tolerating arbitrary failures use significant resources to constantly mask omission and timing failures from faulty processes. This paper presents Quorum Selection, a mechanism that allows to select well functioning processes for active participation in a system. Different from previous work, Quorum Selection not only excludes provably faulty processes that deviated from the protocol, but also takes omission and timing failures into account, even if they only affect individual links. We present a system architecture for Quorum Selection, including a novel failure detector that uses expectations to detect omission and timing failures in a Byzantine environment. We investigate how often an adversary may cause the quorum to changes. We show a quadratic lower bound for general Quorum Selection, but we also define a special case of Quorum Selection for leader based systems that requires trying at most 6f quorums.
HotStuff is a Byzantine fault-tolerant state machine replication protocol that incurs linear communication costs to achieve consensus. This linear scalability promoted the protocol to be adopted as the consensus mechanism in permissioned blockchains. This paper discusses the architecture and evaluation of our extensible framework to implement three HotStuff variants. This reimplementation demonstrates the extensibility of our framework to implement other HotStuff-like protocols. Leveraging our deployment tool, we evaluated our implementation on a wide variety of configurations.