Crust: A Modular Framework for Conflict-free Replicated Data Types (CRDTs) Development, Validation, and Benchmarking

Abstract

Conflict-free Replicated Data Types (CRDTs) are vital for achieving strong eventual consistency in distributed systems, but their development and evaluation face significant challenges by inadequate tooling. While CRDT research focus on algorithms and applications, critical gaps persist in validation and performance benchmarking. To address this, a novel Rust-based framework “Crust” designed to offer a modular, configurable, and extensible platform for developing, validating, and benchmarking CRDT implementations. Crust includes Core, Config, Network, Validation, and Benchmark moduels, supporting various synchronization methods and emphasizing correctness and performance analysis. This paper details Crust’s design and theoretical basis, showing how it bridges the gap between CRDT theory and practice, addressing tooling gaps to enhance adoption and real-world use. By enabling rigorous testing and performance evaluation, Crust has the potential to accelerate the development and adoption of CRDTs in real-world distributed systems.