CNS Core: Small: One Size doesn't Fit All: Workload-Aware Cost Optimization for Decentralized Applications on Blockchains

去中心化应用程序(或Dapp)，如加密货币和其他区块链应用程序，通过不信任可能会受到损害的中央提供商，提供了比传统集中式应用程序更多的好处。因此，Dapp有可能扰乱广泛的业务，特别是分散的金融。然而，由于大量使用昂贵的区块链底层，今天的Dapp给最终用户带来了高昂的货币成本。该项目的目标是在不牺牲安全性或引入额外延迟的情况下，以数量级优化Dapp的成本。PI建议研究之前未开发的设计空间，以实现区块链上的工作负载感知成本优化。PI将设计和评估一系列新的成本节约技术，例如在事务中捆绑请求、动态重组数据存储、引入工作负载自适应控制机制等。一个关键的项目成果将是开发一个成本优化的中间件系统，该系统在实际DAPP工作负载下有效，并且可以很容易地改装到运营区块链(例如，以太)上。还将构建和发布交易记录和回放的软件工具，以实现区块链上的工作负载分析。该项目将提高对动态工作负载下Dapp成本的理解，并影响分布式系统、中间件、区块链和网络安全领域的研究社区。此外，它还将影响各个区块链和DApp开发者社区。它将促进Dapp的采用，并有助于分散许多行业，包括网络、信息技术(IT)和金融技术。最后，PI将利用现有的场所，如MOST(锡拉丘兹科学和技术博物馆)和LSAMP(路易斯·斯托克斯少数群体参与联盟)，有效地接触K-12岁以下学生、代表不足的学生和普通观众。这一奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Decentralized applications (or DApps), such as cryptocurrencies and other blockchain applications, offer benefits over conventional centralized applications by not assuming trust in a central provider, which may be compromised. As a result, DApps have the potential to disrupt a wide range of businesses, notably decentralized finance. However, today's DApps yield a high monetary cost to end-users due to their intensive use of expensive blockchains underneath. This project aims to optimize DApps' costs by orders of magnitude, without sacrificing security or introducing additional delays. The PI proposes to study the previously unexplored design space for workload-aware cost optimization over blockchains. The PI will design and evaluate a wide range of new cost-saving techniques, such as bundling requests in transactions, dynamically reorganizing data storage, introducing workload-adaptive control mechanisms, and others. A key project outcome will be the development of a cost-optimizing middleware system that is effective under real DApp workloads, and that can be easily retrofitted onto operational blockchains (e.g., Ethereum). Software tools for transaction record-and-replay will also be built and released to enable workload analysis on blockchains.This project will improve the understanding of DApps' costs under dynamic workloads and impact the research communities in distributed systems, middleware, blockchains, and cyber-security. In addition, it will impact various blockchain and DApp developer communities. It will facilitate the adoption of DApps and help to decentralize many industries, including the web, information technology (IT), and finance technology. Finally, the PI will leverage existing venues, such as MoST (Museum of Science and Technology in Syracuse) and LSAMP (Louis Stokes Alliance for Minority Participation) for effective outreach to K-12, underrepresented students, and general audiences.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.