Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

NEMO: Faster Parallel Execution for Highly Contended Blockchain Workloads (Full version)

François Ezard, Can Umut Ileri, Jérémie Decouchant

TL;DR

The paper tackles the bottleneck in blockchain execution under high contention by decoupling consensus from execution and optimizing the execution engine. It introduces NEMO, which fuses an object data model with optimistic concurrency control and four innovations: greedy commit for owned-objects, refined dependency tracking, incomplete read/write hints, and dependency-based priority scheduling. In simulated, high-contention workloads, NEMO delivers up to 42% higher throughput than Block-STM and up to 61% higher than a PCC baseline, while reducing redundant re-executions. These results show that leveraging object-level state and partial prior knowledge can significantly improve blockchain throughput, especially as contention rises.

Abstract

Following the design of more efficient blockchain consensus algorithms, the execution layer has emerged as the new performance bottleneck of blockchains, especially under high contention. Current parallel execution frameworks either rely on optimistic concurrency control (OCC) or on pessimistic concurrency control (PCC), both of which see their performance decrease when workloads are highly contended, albeit for different reasons. In this work, we present NEMO, a new blockchain execution engine that combines OCC with the object data model to address this challenge. NEMO introduces four core innovations: (i) a greedy commit rule for transactions using only owned objects; (ii) refined handling of dependencies to reduce re-executions; (iii) the use of incomplete but statically derivable read/write hints to guide execution; and (iv) a priority-based scheduler that favors transactions that unblock others. Through simulated execution experiments, we demonstrate that NEMO significantly reduces redundant computation and achieves higher throughput than representative approaches. For example, with 16 workers NEMO's throughput is up to 42% higher than the one of Block-STM, the state-of-the-art OCC approach, and 61% higher than the pessimistic concurrency control baseline used.

NEMO: Faster Parallel Execution for Highly Contended Blockchain Workloads (Full version)

TL;DR

The paper tackles the bottleneck in blockchain execution under high contention by decoupling consensus from execution and optimizing the execution engine. It introduces NEMO, which fuses an object data model with optimistic concurrency control and four innovations: greedy commit for owned-objects, refined dependency tracking, incomplete read/write hints, and dependency-based priority scheduling. In simulated, high-contention workloads, NEMO delivers up to 42% higher throughput than Block-STM and up to 61% higher than a PCC baseline, while reducing redundant re-executions. These results show that leveraging object-level state and partial prior knowledge can significantly improve blockchain throughput, especially as contention rises.

Abstract

Following the design of more efficient blockchain consensus algorithms, the execution layer has emerged as the new performance bottleneck of blockchains, especially under high contention. Current parallel execution frameworks either rely on optimistic concurrency control (OCC) or on pessimistic concurrency control (PCC), both of which see their performance decrease when workloads are highly contended, albeit for different reasons. In this work, we present NEMO, a new blockchain execution engine that combines OCC with the object data model to address this challenge. NEMO introduces four core innovations: (i) a greedy commit rule for transactions using only owned objects; (ii) refined handling of dependencies to reduce re-executions; (iii) the use of incomplete but statically derivable read/write hints to guide execution; and (iv) a priority-based scheduler that favors transactions that unblock others. Through simulated execution experiments, we demonstrate that NEMO significantly reduces redundant computation and achieves higher throughput than representative approaches. For example, with 16 workers NEMO's throughput is up to 42% higher than the one of Block-STM, the state-of-the-art OCC approach, and 61% higher than the pessimistic concurrency control baseline used.
Paper Structure (15 sections, 10 figures, 1 table)

This paper contains 15 sections, 10 figures, 1 table.

Table of Contents

  1. Introduction
  2. Related Work
  3. Background
  4. NEMO
  5. Combining the object-model and OCC
  6. Greedy commit rule
  7. Limiting re-executions
  8. Incomplete hints
  9. Performance Evaluation
  10. Implementation
  11. Workloads
  12. Number of re-executions
  13. Throughput
  14. Discussion and Future Work
  15. Conclusion

Figures (10)

  • Figure 1: Block-STM's architecture
  • Figure 2: Transaction states and transitions between them in Block-STM
  • Figure 3: Transaction lifecycle (from SuiLutris)
  • Figure 4: Example transaction logic showing how prior knowledge differs: exhaustive knowledge includes all possible writes ($A$, $B$, $C$), while partial knowledge only includes guaranteed writes ($A$).
  • Figure : (a) 8 Workers
  • ...and 5 more figures