III: Small: Peer-to-peer Database (P2PDB): A decentralized, scalable data sharing and management platform

III：小型：点对点数据库（P2PDB）：去中心化、可扩展的数据共享和管理平台

• 批准号: 1910613
• 负责人: Daniel Abadi
• 金额: $ 50万
• 依托单位: University of Maryland, College Park
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2019
• 资助国家: 美国
• 起止时间: 2019-10-01 至 2023-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1910613&HistoricalAwards=false
• 关键词: III; Small; Peer; peer; Database

This project involves designing and implementing a new system, called Peer-to-peer Database (P2PDB), that will serve as a decentralized platform for data publishing, sharing, and querying of data. P2PDB enables an unlimited number of independent participants to publish and access the contents of datasets stored across the participants. The system will provide decentralized publishing and querying functionality over structured data in an analogous fashion to how the "world wide web" enables decentralized publishing and accessing of unstructured data. However, P2PDB differs from both the traditional world wide web (and also other attempts at building peer-to-peer structured database systems) in the way that it provides incentives for performing tasks that are critical to the well-being of the system as a whole, including contribution, integration, storing, and processing of data, and enforcing good behavior by the participants. This project includes involvement in the research track at "Technica," the largest all-female hackathon in the US hosted every year at the University of Maryland, and computer science curricular development at the local high school level.The project includes working on several challenging research problems that result from an architectural assumption that any entity may participate in the storage and processing of data in P2PDB. First, the system must be robust in the face of malicious nodes that join the system and intentionally corrupt data or return incorrect results upon receiving a query. This requires research effort in the areas of query verification, secure data processing, and decentralized dispute moderation. Another important research problem that emerges from the planned architecture of P2PDB is query optimization and execution over a federated architecture in which queries must be performed across many different entities, each with different service level guarantees and costs.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

该项目涉及设计和实施一个名为对等数据库（P2PDB）的新系统，该系统将作为数据发布、共享和查询的分散平台。P2PDB允许无限数量的独立参与者发布和访问存储在参与者之间的数据集的内容。该系统将以类似于“万维网”如何使非结构化数据能够分散发布和访问的方式，提供结构化数据的分散发布和查询功能。然而，P2PDB与传统的万维网（以及其他构建对等结构化数据库系统的尝试）的不同之处在于，它为执行对整个系统的健康至关重要的任务提供了激励，包括贡献，集成，存储和处理数据，以及参与者的良好行为。该项目包括参与“Technica”的研究轨道，该项目是美国马里兰州大学每年举办的最大的全女性黑客活动，以及当地高中水平的计算机科学课程开发。该项目包括研究几个具有挑战性的研究问题，这些问题源于一个架构假设，即任何实体都可以参与P2PDB中数据的存储和处理。首先，系统必须在面对加入系统并故意损坏数据或在接收查询时返回错误结果的恶意节点时保持健壮。这需要在查询验证、安全数据处理和分散式争议调解等领域进行研究。另一个重要的研究问题，出现从计划的P2PDB架构是查询优化和执行的联邦体系结构中，查询必须在许多不同的实体，每个不同的服务水平保证和costs.This奖项反映了NSF的法定使命，并已被认为是值得的支持，通过评估使用基金会的智力价值和更广泛的影响审查标准。

项目成果

C5: Cloned Concurrency Control That Always Keeps Up

• DOI: 10.14778/3561261.3561262
• 发表时间: 2022-07
• 期刊: Proc. VLDB Endow.
• 作者: Jeffrey Helt;Abhinav Sharma;D. Abadi;Wyatt Lloyd;Jose M. Faleiro

BullFrog: Online Schema Evolution via Lazy Evaluation

• DOI: 10.1145/3448016.3452842
• 发表时间: 2021-06
• 期刊: Proceedings of the 2021 International Conference on Management of Data
• 作者: Souvik Bhattacherjee;Gang Liao;M. Hicks;D. Abadi

Integration of Large-Scale Data Processing Systems and Traditional Parallel Database Technology

• DOI: 10.14778/3352063.3352145
• 发表时间: 2019-08
• 期刊: Proc. VLDB Endow.
• 作者: A. Abouzeid;D. Abadi;Kamil Bajda-Pawlikowski;A. Silberschatz

FileScale: Fast and Elastic Metadata Management for Distributed File Systems

• DOI: 10.1145/3620678.3624784
• 发表时间: 2023-10
• 期刊: Proceedings of the 2023 ACM Symposium on Cloud Computing
• 作者: Gangjun Liao;Daniel J. Abadi

Detock: High Performance Multi-region Transactions at Scale

• DOI: 10.1145/3589293
• 发表时间: 2023-06
• 期刊: Proceedings of the ACM on Management of Data
• 作者: Cuong D. T. Nguyen;Johann K. Miller;D. Abadi