Back to the Future: Rethinking Endorsement in Order-Execute Blockchains

Rongji Huang, Yifeng Ye, Gerui Wang, Mingchao Wan, Yuxing Duan, Jingjing Zhang, Guangtao Xue, Shengyun Liu · 2026

Read Back to the Future: Rethinking Endorsement in Order-Execute Blockchains by Rongji Huang, Yifeng Ye, Gerui Wang, Mingchao Wan, Yuxing Duan, Jingjing Zhang, Guangtao Xue, Shengyun Liu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-27659.txt.

Introduction

Huang and seven colleagues bring flexible endorsement to order-execute blockchains with FlexTender, built on Tendermint. Modern private blockchains need endorsement rules defined per contract or state object, which Fabric handles with execute-order-validate, while most systems, including ChainMaker, use order-execute. It embeds them into consensus, removing problem transactions from an ordered list deterministically, keeping censorship resistance and decentralization intact. On an Ethereum USDT workload, it reaches up to 10.6 times the throughput of an execute-order-validate test on the same platform.

Catalog record

CREATORRongji Huang, Yifeng Ye, Gerui Wang, Mingchao Wan, Yuxing Duan, Jingjing Zhang, Guangtao Xue, Shengyun Liu
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE83376
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.27659

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Back to the Future: Rethinking Endorsement in Order-Execute Blockchains Rongji Huang, Yifeng Ye, Gerui Wang, Mingchao Wan, Yuxing Duan, Jingjing Zhang, Guangtao Xue, Shengyun Liu Abstract Due to regulatory compliance and governance management, modern (permissioned) blockchains require flexible endorsement, which allows the endorsement policy for each contract or state object to be individually defined. To enable flexible endorsement, Hyperledger Fabric employs an execute-order-validate (EOV) paradigm, in which transactions first undergo speculative execution and endorsement, and are only then ordered and validated. Meanwhile, most blockchain systems, including the platform targeted in this work (i.e., ChainMaker), still follow a conflict-free order-execute framework. We argue that the EOV paradigm still faces several limitations, notably high abort rates in high-contention workloads such as those in Decentralized Finance (DeFi). To avoid refactoring our system and better suit DeFi applications, we try to integrate flexible endorsement into the classical order-execute architecture and accordingly propose a new framework. The key challenge is to deterministically remove problematic transactions from an ordered list, while preserving censorship resistance and decentralization for the remaining ones. We instantiate this framework on top of Tendermint, a seminal Byzantine fault-tolerant (BFT) protocol adopted in our system, and thereby propose FlexTender . By elegantly embedding endorsements into consensus, FlexTender incurs no additional messaging overhead in the normal case. Empirical evaluation using an Ethereum USDT workload demonstrates that FlexTender achieves up to 10.6 × 10.6\times speedup in throughput over an EOV simulation on the same platform. 1 Introduction Decentralization is fundamental to the prosperity of blockchain ecosystems [ 85 , 9 , 83 ] . However, this advantage comes with an inevitable cost: the technology can be easily exploited for illicit activities [ 27 , 86 , 84 ] . Even without malicious intent, accidents [ 62 , 80 , 56 ] can also erode the trust that underpins blockchain systems. With the recent establishment of regulatory frameworks (e.g., the U.S. GENIUS Act [ 66 ] and the Hong Kong Stablecoins Ordinance [ 82 ] ), it has become more evident that regulation is essential for the widespread adoption of decentralized systems. In this work, we aim to integrate proactive regulation into ChainMaker [ 21 ] , an enterprise-grade blockchain platform developed in part by some of the authors and rolled out across dozens of state-owned enterprises in China. Toward this goal, we found that a key step is to support heterogeneous roles in the validation of specific transactions and their execution results. A similar idea, also referred to as flexible endorsement , was first introduced in Hyperledger Fabric [ 5 ] (hereafter, Fabric). To support customizable endorsement policies and overcome limitations inherent in the traditional state machine replication (SMR) paradigm [ 71 ] , Fabric adopts a novel execute-order-validate (EOV) framework inspired by Eve [ 45 ] . In Fabric, transactions are first simulated (i.e., pre-executed) in parallel by endorsers 1 1 1 In Fabric, endorsers are nodes that maintain local copies of the ledger and are responsible for executing and endorsing transactions. , producing read/write sets. Compliant transactions and their execution results are then endorsed (through signatures). Based on predefined endorsement policies, an ordering service sequences properly endorsed transactions, typically via a consensus protocol. As transactions were pre-executed in parallel, a final validation phase checks for read-write conflicts and aborts conflicting transactions [ 1 ] . To be committed, aborted transactions must be retried. Despite Fabric’s appealing features and proven enterprise success, the classical order-execute framework continues to dominate blockchain systems [ 16 , 39 , 52 , 57 , 17 , 33 ] . While workarounds exist for most limitations addressed by Fabric, supporting flexible endorsement presents inherent difficulties. Most notably, as endorsement policies may implement arbitrary logic, whether a transaction can be properly endorsed goes beyond a consensus problem. A key benefit of the EOV paradigm is its simplicity in handling endorsement, as any transaction remains off the critical path of others until getting endorsed 2 2 2 This is likely the main reason why Fabric adopted an “inverted” structure. . Nonetheless, placing endorsement ahead still entails several limitations, rendering it unsuitable for modern order-execute blockchains. • Fabric inverts the order-execute model and explicitly relies on a validation phase to ensure serializability; Its architectural idea does not transfer readily to order-execute blockchains, as it affects not only consensus but also execution semantics; • As there is no sequencing in the execution phase, each transaction is endorsed individually, which hinders batching—a key technique for amortizing the cost of cryptographic operations [ 79 ] for endorsement; and, • The EOV framework is ill-suited for high-contention workloads, which are prevalent in financial applications like decentralized exchanges (DEX) [ 85 ] ; The problem not only degrades performance but may undermine liveness when certain transactions accessing hot-spot state are repeatedly aborted (see Figure 12 ). Numerous improvements [ 35 , 60 , 10 , 58 , 70 , 36 , 73 , 75 , 81 ] have been made to Fabric, yet they remain constrained by the original framework and inherited limitations. We instead pivot toward the alternative paradigm. We observe that the key to enabling certain participants to play critical roles in validating specific transactions is to keep the rest of the system decentralized and as unaffected as possible 3 3 3 In principle, endorsement policies should themselves be (partially) decentralized, although this is ultimately the responsibility of application designers. ; Otherwise, one might simply adopt a centralized approach. Targeting predominant order-execute blockchains, we study a framework termed order-execute-endorse , in which endorsement occurs only after transactions have been ordered and executed. The rationale behind our design is simple: Rather than simplifying endorsement handling at the cost of refactoring and introducing conflicts, our framework aligns with most existing systems and remains conflict-free. Challenges. We however must deal with the cases where some transactions fail to obtain endorsements. Since such problematic transactions may affect the outcomes of subsequent ones, nodes must remove or eventually commit them in a deterministic manner. More importantly, subsequent transactions should be re-executed and re-endorsed. The cycle of removal and re-execution may repeat iteratively until all remaining transactions are endorsed. As transactions are removable, we should also address the issue of endorsable transactions being deliberately removed by malicious nodes (i.e., censorship attacks). It is worth noting that early versions [ 8 ] of Fabric (i.e., v0.6 [ 43 ] and earlier) were built on the order-execute paradigm, yet they lacked support for flexible endorsement. This architectural shift highlights the difficulty of introducing flexible endorsement into an order-execute blockchain. We initially attempted to introduce an endorsement phase as a standalone component, similar to Fabric’s modular architecture, but found the resulting design too complicated for practical implementation and real-world deployment. As transaction removal and re-execution may be performed iteratively, such a modular design gives rise to nested rounds of processing: one for consensus (e.g., view changes) and the other for endorsement. We observe that the need to remove transactions and reprocess the remaining ones bears similarities with the need to advance consensus in a round-by-round fashion. We thus try to elegantly integrate flexible endorsement into consensus messages. To this end, • The round (or view) advancement not merely facilitates leader election, but also aligns

Read the complete text