Sharding SMR with Optimal-size Shards for Highly Scalable Blockchains

Sharding can enhance blockchain scalability by dividing nodes into multiple shards to handle transactions in parallel. However, the size-security dilemma where a shard must be large enough to ensure its security constrains the overall number of shards, rendering blockchain sharding low parallelism and poor scalability. This paper presents Arete, an optimally scalable blockchain sharding architecture designed to resolve the dilemma based on a key observation: higher (Byzantine) fault-resilient shards allow the creation of more secure shards. The main idea of Arete, therefore, is to improve the security resilience/threshold of shards by sharding the blockchain's State Machine Replication (SMR) process itself. First, Arete decouples the three steps in SMR, leading to a single ordering shard performing the ordering task and multiple processing shards performing the dispersing and execution tasks. This frees processing shards from running consensus, allowing up to half compromised nodes per processing shard. Second, Arete considers safety and liveness against Byzantine failures separately to improve the safety threshold further while tolerating temporary liveness violations in a controlled manner. Apart from the creation of more optimal-size shards, such a deconstructed SMR scheme also empowers us to devise a novel certify-order-execute model to fully parallelize transaction handling, thereby significantly improving the performance of sharded blockchain systems. We implement Arete and evaluate it on a geo-distributed AWS environment. Our results demonstrate that Arete outperforms the state-of-the-art sharding protocol in terms of transaction throughput and cross-shard confirmation latency without compromising on intra-shard confirmation latency.