Scaling blockchain via sharding

Abstract

As a promising solution to blockchain scalability, sharding divides blockchain nodes into small groups called shards, splitting the workload. Existing works for sharding, however, are limited by three challenges. First, cross-shard transactions multiply the overhead of blockchain sharding, since the system needs to split each cross-shard transaction into multiple sub-transactions, each of which costs a consensus round to commit. Second, the independent and random scheduling for cross-shard transactions in different shards results in numerous conflicts and aborts since their involved accounts may be modified or locked by the others before they are committed. Third, for a blockchain database, a new and popular blockchain application, it is challenging to construct a scalable blockchain database through traditional on-chain sharding. Therefore, in chapter 3, we present a new blockchain sharding schema, named layered sharding, to improve the scalability via sharding while processing the cross-shard transactions efficiently. In chapter 4, we present PROPHET, a sharding blockchain system with deterministic ordering for conflict-free transactions. In chapter 5, we propose GrIDB, the first scalable blockchain database that distributes tables to different shards and provides efficient cross-shard database services. Evaluation of the real-world datasets for blockchain shows the remarkable performance improvement of our proposed systems over existing solutions.