Untangling Blockchain: A Data Processing View of Blockchain Systems (1708.05665v1)

Abstract: Blockchain technologies are gaining massive momentum in the last few years. Blockchains are distributed ledgers that enable parties who do not fully trust each other to maintain a set of global states. The parties agree on the existence, values and histories of the states. As the technology landscape is expanding rapidly, it is both important and challenging to have a firm grasp of what the core technologies have to offer, especially with respect to their data processing capabilities. In this paper, we first survey the state of the art, focusing on private blockchains (in which parties are authenticated). We analyze both in-production and research systems in four dimensions: distributed ledger, cryptography, consensus protocol and smart contract. We then present BLOCKBENCH, a benchmarking framework for understanding performance of private blockchains against data processing workloads. We conduct a comprehensive evaluation of three major blockchain systems based on BLOCKBENCH, namely Ethereum, Parity and Hyperledger Fabric. The results demonstrate several trade-offs in the design space, as well as big performance gaps between blockchain and database systems. Drawing from design principles of database systems, we discuss several research directions for bringing blockchain performance closer to the realm of databases.

• The paper’s main contribution is the BLOCKBENCH framework, which benchmarks blockchain performance on data processing workloads.
• The study categorizes blockchain systems by distributed ledger, consensus protocol, cryptography, and smart contracts to clarify their strengths and limitations.
• The evaluation reveals that while private blockchains like Hyperledger achieve higher throughput, they face scalability challenges compared to public blockchains.

Untangling Blockchain: A Data Processing View of Blockchain Systems

Blockchain technologies have rapidly emerged, finding application in a myriad of domains beyond their initial use in crypto-currencies. The paper "Untangling Blockchain: A Data Processing View of Blockchain Systems" addresses the core challenge of comprehensively understanding blockchain systems, specifically within the field of data processing. This essay provides an overview of the paper, emphasizing its insights on blockchain technology, evaluation of prominent blockchain systems, and avenues for future research.

Overview

The authors categorize blockchain systems using four key concepts: distributed ledger, consensus protocol, cryptography, and smart contracts. They conduct an in-depth survey of both in-production and research blockchain systems, with a primary focus on private blockchains where participants are authenticated. Additionally, the paper introduces BLOCKBENCH, a benchmarking framework aimed at evaluating blockchain performance against data processing workloads.

Key Contributions

1. Categorization of Blockchain Systems: The paper distinguishes between public and private blockchains and breaks down the technical aspects into distributed ledger, consensus protocol, cryptography, and smart contracts. This taxonomy is instrumental in understanding how different blockchain systems operate and their inherent strengths and limitations.
2. BLOCKBENCH Benchmarking Framework: BLOCKBENCH is presented as a versatile tool for benchmarking blockchain systems. It includes micro and macro benchmarks, exposes blockchain performance bottlenecks, and compares Ethereum, Parity, and Hyperledger.
3. Evaluation of Blockchain Systems: Using BLOCKBENCH, the paper provides a comprehensive evaluation of Ethereum, Parity, and Hyperledger. The results show that while Hyperledger generally performs better across various benchmarks, both Ethereum and Parity exhibit resilience to node failures but at the cost of potential security vulnerabilities like fork attacks.
4. Lessons from Comparative Analysis:

The paper draws several critical insights: - Blockchain systems are not yet on par with traditional database systems in terms of data processing performance. - There are significant opportunities for performance improvements by adopting principles from database systems. - Trusted hardware and sharding are promising areas for enhancing blockchain scalability and efficiency.

Detailed Analysis

Distributed Ledger

The structure of the distributed ledger in blockchain systems can vary significantly. Public blockchains like Bitcoin and Ethereum use a global ledger model, where any participant can join and update the ledger. In contrast, private blockchains like Hyperledger often utilize a permissioned model with controlled access, making them more suitable for enterprise applications requiring authenticated and restricted interactions.

Consensus Protocol

The consensus protocol is crucial in ensuring that all nodes in a blockchain system agree on the state of the ledger. Public blockchains typically rely on Proof-of-Work (PoW), which although secure, is computationally intensive and slow. Private blockchains may use more efficient consensus mechanisms like Practical Byzantine Fault Tolerance (PBFT) or Proof-of-Authority (PoA), both of which are more performant but assume a level of trust among participants.

Cryptographic Techniques

Blockchains heavily rely on cryptographic methods to ensure data integrity and authenticity. Techniques such as Merkle trees and cryptographic hash functions are standard for ensuring the immutability of the ledger. Additionally, the use of public key infrastructure (PKI) for identity management and transaction validation is crucial for the security of blockchain systems.

Smart Contracts

Smart contracts extend blockchain functionality by allowing the execution of user-defined scripts or programs on the blockchain. There is a spectrum of expressiveness in smart contract languages, ranging from the constrained script languages used in Bitcoin to the Turing-complete languages used in Ethereum. The latter enables more complex applications but comes with increased risks of security vulnerabilities, as evidenced by incidents like the DAO attack on Ethereum.

Evaluation with BLOCKBENCH

The evaluation conducted with BLOCKBENCH highlights several critical performance characteristics:

• Throughput and Latency: Hyperledger consistently achieves higher throughput and lower latency compared to Ethereum and Parity. This can be attributed to its use of PBFT, which, despite its limitations in scalability, provides efficient consensus under smaller network sizes.
• Scalability: Ethereum and Parity exhibit better scalability in larger networks compared to Hyperledger, which struggles beyond 16 nodes due to network message congestion.
• Fault Tolerance and Security: Ethereum and Parity show vulnerability to network partition attacks, leading to blockchain forks. Hyperledger, owing to its PBFT consensus, maintains safety but at the cost of higher recovery time after network disruptions.

Implications and Future Research

The paper’s findings present significant implications for both theoretical and practical advancements in blockchain technology:

1. Integration of Database Principles: Drawing design principles from traditional database systems can help mitigate some of the performance bottlenecks in blockchain systems. For instance, sharding and optimized data models can significantly enhance throughput and reduce latencies.
2. Leveraging New Hardware: Employing trusted hardware like Intel SGX can streamline consensus by reducing the overhead associated with Byzantine fault tolerance. This shift can lead to more efficient and scalable blockchain architectures.
3. Enhanced Data Models and Smart Contract Languages: Advanced data models that support fine-grained versioning and declarative smart contract languages can improve both performance and usability. Ensuring that these models and languages can be formally verified will also enhance security and reliability.
4. Benchmarking and Standardization: Continued development and adoption of frameworks like BLOCKBENCH are essential for standardizing the performance evaluation of blockchain systems. This will enable more objective comparisons and drive improvements in blockchain technology.

Conclusion

The paper "Untangling Blockchain: A Data Processing View of Blockchain Systems" provides a thorough examination of blockchain systems from a data processing perspective. By categorizing existing systems, introducing the BLOCKBENCH benchmarking framework, and evaluating major blockchain platforms, the authors lay the groundwork for ongoing research and development in blockchain technology. Future work should focus on integrating database optimization techniques, leveraging new hardware capabilities, and enhancing the programmability and security of smart contracts. This research is critical for advancing the scalability, performance, and widespread adoption of blockchain systems in diverse application domains.