TY - GEN
T1 - BFT-Store
T2 - 36th IEEE International Conference on Data Engineering, ICDE 2020
AU - Qi, Xiaodong
AU - Zhang, Zhao
AU - Jin, Cheqing
AU - Zhou, Aoying
N1 - Publisher Copyright:
© 2020 IEEE.
PY - 2020/4
Y1 - 2020/4
N2 - The full-replication data storage mechanism, as commonly utilized in existing blockchain systems, is lack of sufficient storage scalability, since it reserves a copy of the whole block data in each node so that the overall storage consumption per block is O(n) with n nodes. Moreover, due to the existence of Byzantine nodes, existing partitioning methods, though widely adopted in distributed systems for decades, cannot suit for blockchain systems directly, thereby it is critical to devise a new storage mechanism. This paper proposes a novel storage engine, called BFT-Store, to enhance storage scalability by integrating erasure coding with Byzantine Fault Tolerance (BFT) consensus protocol. First, the storage consumption per block can be reduced to O(1), which enlarges overall storage capability when more nodes join blockchain. Second, an efficient online re-encoding protocol is designed for storage scale-out and a hybrid replication scheme is employed to improve reading performance. Last, extensive experimental results illustrate the scalability, availability and efficiency of BFT-Store, which is implemented on an open-source permissioned blockchain Tendermint.
AB - The full-replication data storage mechanism, as commonly utilized in existing blockchain systems, is lack of sufficient storage scalability, since it reserves a copy of the whole block data in each node so that the overall storage consumption per block is O(n) with n nodes. Moreover, due to the existence of Byzantine nodes, existing partitioning methods, though widely adopted in distributed systems for decades, cannot suit for blockchain systems directly, thereby it is critical to devise a new storage mechanism. This paper proposes a novel storage engine, called BFT-Store, to enhance storage scalability by integrating erasure coding with Byzantine Fault Tolerance (BFT) consensus protocol. First, the storage consumption per block can be reduced to O(1), which enlarges overall storage capability when more nodes join blockchain. Second, an efficient online re-encoding protocol is designed for storage scale-out and a hybrid replication scheme is employed to improve reading performance. Last, extensive experimental results illustrate the scalability, availability and efficiency of BFT-Store, which is implemented on an open-source permissioned blockchain Tendermint.
KW - Blockchain
KW - Byzantine Fault Tolerance
KW - Erasure coding
KW - Storage scalability
UR - https://www.scopus.com/pages/publications/85085859553
U2 - 10.1109/ICDE48307.2020.00205
DO - 10.1109/ICDE48307.2020.00205
M3 - 会议稿件
AN - SCOPUS:85085859553
T3 - Proceedings - International Conference on Data Engineering
SP - 1926
EP - 1929
BT - Proceedings - 2020 IEEE 36th International Conference on Data Engineering, ICDE 2020
PB - IEEE Computer Society
Y2 - 20 April 2020 through 24 April 2020
ER -