Ira: Efficient Transaction Replay for Distributed Systems

2026-01-29

AI Summary

This work addresses the performance bottleneck in primary-backup replication caused by transaction replay latency at backup nodes. The authors propose Ira-L, a novel protocol that leverages the primary node’s knowledge of future data access patterns to generate lightweight hints—comprising key working sets and table metadata—which are sent alongside transaction batches to guide multi-threaded prefetching and caching optimizations at backups. Crucially, Ira-L achieves substantial replay acceleration without altering the underlying protocol semantics. Evaluation using real-world Ethereum mainnet data demonstrates that Ira-L improves the median per-block replay speed at backup nodes by 25×. With 16 prefetching threads, total replay time for the evaluated workload drops from 6.5 hours to just 16 minutes, with hint transmission overhead amounting to only approximately 5%.

Technology Category

• Search and Optimization: Distributed Search
• Planning, Routing, and Scheduling: Replanning and Plan Repair
• Multiagent Systems: Mechanism Design

Application Category

• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Security and Privacy: Privacy-enhancing technologies

Abstract

In primary-backup replication, consensus latency is bounded by the time for backup nodes to replay (re-execute) transactions proposed by the primary. In this work, we present Ira, a framework to accelerate backup replay by transmitting compact \emph{hints} alongside transaction batches. Our key insight is that the primary, having already executed transactions, possesses knowledge of future access patterns which is exactly the information needed for optimal replay. We use Ethereum for our case study and present a concrete protocol, Ira-L, within our framework to improve cache management of Ethereum block execution. The primaries implementing Ira-L provide hints that consist of the working set of keys used in an Ethereum block and one byte of metadata per key indicating the table to read from, and backups use these hints for efficient block replay. We evaluated Ira-L against the state-of-the-art Ethereum client reth over two weeks of Ethereum mainnet activity ($100,800$ blocks containing over $24$ million transactions). Our hints are compact, adding a median of $47$ KB compressed per block ($\sim5\%$ of block payload). We observe that the sequential hint generation and block execution imposes a $28.6\%$ wall-time overhead on the primary, though the direct cost from hints is $10.9\%$ of execution time; all of which can be pipelined and parallelized in production deployments. On the backup side, we observe that Ira-L achieves a median per-block speedup of $25\times$ over baseline reth. With $16$ prefetch threads, aggregate replay time drops from $6.5$ hours to $16$ minutes ($23.6\times$ wall-time speedup).

Problem

• transaction replay
• distributed systems
• primary-backup replication
• consensus latency
• backup execution

Innovation

• transaction replay
• hint-based prefetching
• primary-backup replication
• Ethereum execution
• distributed consensus acceleration