Ladon: High-Performance Multi-BFT Consensus via Dynamic Global Ordering (Extended Version)

TL;DR

Ladon addresses the bottleneck of global ordering in Multi-BFT consensus by introducing dynamic global ordering with monotonic ranks, enabling blocks from parallel instances to be ordered according to their real-time production while preserving inter-block causality. The approach decouples partial logs, pipelines rank coordination with consensus, and uses aggregate signatures to reduce overhead, allowing Ladon to integrate with PBFT or HotStuff. Empirical evaluation on WAN/LAN shows Ladon achieving up to approximately 8x throughput and substantial latency reductions in the presence of stragglers, with robust causality guarantees and resilience to Byzantine behavior. The work demonstrates that dynamic, rank-based global ordering can markedly improve scalability and reliability of large-scale distributed ledgers and fault-tolerant systems.

Abstract

Multi-BFT consensus runs multiple leader-based consensus instances in parallel, circumventing the leader bottleneck of a single instance. However, it contains an Achilles' heel: the need to globally order output blocks across instances. Deriving this global ordering is challenging because it must cope with different rates at which blocks are produced by instances. Prior Multi-BFT designs assign each block a global index before creation, leading to poor performance. We propose Ladon, a high-performance Multi-BFT protocol that allows varying instance block rates. Our key idea is to order blocks across instances dynamically, which eliminates blocking on slow instances. We achieve dynamic global ordering by assigning monotonic ranks to blocks. We pipeline rank coordination with the consensus process to reduce protocol overhead and combine aggregate signatures with rank information to reduce message complexity. Ladon's dynamic ordering enables blocks to be globally ordered according to their generation, which respects inter-block causality. We implemented and evaluated Ladon by integrating it with both PBFT and HotStuff protocols. Our evaluation shows that Ladon-PBFT (resp., Ladon-HotStuff) improves the peak throughput of the prior art by $\approx$8x (resp., 2x) and reduces latency by $\approx$62% (resp., 23%), when deployed with one straggling replica (out of 128 replicas) in a WAN setting.

Figures (10)

• Figure 1: An overview of Multi-BFT paradigm. The shaded (resp., white) blocks refer to the globally confirmed (resp., partially committed) blocks. The dashed blocks refer to the blocks to be produced in the future.
• Figure 2: The analytical and experimental performance of Multi-BFT consensus with/without a straggler. The vertical line in (a) represents the queued partially committed blocks, while the horizontal line represents the delay of the global ordering.
• Figure 3: An illustration of the dynamic global ordering process. At time $t_1$, a new block $B_2^2$ is partially committed, which makes blocks $B_2^1$ and $B_3^0$ globally confirmed.
• Figure 4: An overview of Ladon's design. Client transactions are enqueued into rotating buckets and are then consumed by concurrent BFT protocol instances. Each instance packs transactions into blocks that are eventually totally ordered. Instances execute in epochs.
• Figure 5: Throughput and latency of Ladon, ISS, RCC, Mir, and DQBFT in WAN (a) -- (d), and LAN (e) -- (h).

Theorems & Definitions (10)

• Lemma 1
• Theorem 1: MR-Agreement
• Theorem 2: MR-Monotonicity
• Lemma 2
• Lemma 3
• Theorem 3: Totality
• Lemma 4
• Theorem 4: Agreement
• Lemma 5
• Theorem 5: Liveness