Orthrus: Accelerating Multi-BFT Consensus through Concurrent Partial Ordering of Transactions (Extended Version)

TL;DR

Orthrus tackles the inherent latency bottleneck of Multi-BFT consensus by enabling partial ordering for conflict-free payments and hybrid global ordering for contract transactions, coordinated through an escrow-based atomicity mechanism. The approach partitions transactions into per-instance buckets, uses sequenced broadcast to form partial and global logs, and executes payments concurrently while contracting operations run under a global order; an escrow layer ensures atomicity across multiple instances. The authors provide a formal model, correctness proofs, and extensive WAN/LAN experiments on up to 128 replicas, showing significant latency reductions (up to 87% in WAN) and robust performance under stragglers and faults. The results indicate practical impact for large-scale, cross-region deployments, offering substantial improvements in end-to-end latency and throughput for mixed workloads that include both payments and smart-contract-style transactions.

Abstract

Multi-Byzantine Fault Tolerant (Multi-BFT) consensus allows multiple consensus instances to run in parallel, resolving the leader bottleneck problem inherent in classic BFT consensus. However, the global ordering of Multi-BFT consensus enforces a strict serialized sequence of transactions, imposing additional confirmation latency and also limiting concurrency. In this paper, we introduce Orthrus, a Multi-BFT protocol that accelerates transaction confirmation through partial ordering while reserving global ordering for transactions requiring stricter sequencing. To this end, Orthrus strategically partitions transactions to maximize concurrency and ensure consistency. Additionally, it incorporates an escrow mechanism to manage interactions between partially and globally ordered transactions. We evaluated Orthrus through extensive experiments in realistic settings, deploying 128 replicas in WAN and LAN environments. Our findings demonstrate latency reductions of up to 87% in WAN compared to existing Multi-BFT protocols.

Figures (9)

• Figure 1: (a) Multi-BFT paradigm. The $j$th block produced by instance $i$ is denoted as $B_j^i$. (b) Breakdown latency with a straggler. The green bar refers to transaction transmission delay (① and ④), the orange refers to the delay of a block being delivered from consensus (②), and the black refers to the global ordering delay (③).
• Figure 2: An overview of Orthrus. Transactions are partitioned into distinct buckets, ordered according to their types, and finally executed.
• Figure 3: Throughput and latency of Orthrus, ISS, RCC, Mir, DQBFT, and Ladon in WAN.
• Figure 4: Throughput and latency of Orthrus, ISS, RCC, Mir, DQBFT, and Ladon in LAN.
• Figure 5: Throughput and latency of Orthrus under different payment transactions proportions in WAN.

Theorems & Definitions (16)

• Lemma 1
• proof
• Lemma 2
• proof
• Lemma 3
• proof
• Theorem 1: Safety
• proof
• Lemma 4
• proof