EAGER: Exploring the Feasibility of System Support for Managing Risk in Cloud Markets

EAGER：探索云市场风险管理系统支持的可行性

• 批准号: 1802523
• 负责人: David Irwin
• 金额: $ 30万
• 依托单位: University of Massachusetts Amherst
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2018
• 资助国家: 美国
• 起止时间: 2018-05-01 至 2021-04-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1802523&HistoricalAwards=false
• 关键词: EAGER; Exploring; Feasibility; System; Support

Cloud computing has become the foundation of our information-based economy, providing the computing power necessary for advances in transportation, energy, communication, healthcare, science, and entertainment. Early cloud platforms were simple, requiring users to choose from a small number of servers "on-demand" for a fixed per-hour price. However, modern cloud platforms are evolving into full-fledged markets that offer servers under a range of complex contracts that expose applications to different forms of risk, such as the risk of price increases, server revocations, or server shortages. By offering such contracts, cloud platforms can increase their infrastructure's utilization, revenue, and energy-efficiency, while cloud users can lower their costs by tailoring their contracts to match their applications' requirements. Unfortunately, given the complexity of cloud contracts and the relationship between risk, application correctness, and performance, developing risk-aware applications represents a significant challenge. To address the challenge, this project will explore the feasibility of system support for managing risk in cloud markets by adapting and extending concepts from economics and finance. A key goal of the project is to elevate risk management to a first-class systems design principle. The project includes multiple broader impacts. By enabling better risk management, cloud platforms can significantly increase their utilization and energy-efficiency, while also lowering users' costs. The project will incorporate modules on cloud economics and managing risk in cloud platforms into undergraduate and graduate seminars, which will relate it to core concepts in operating systems, networking, and distributed systems. The researchers also plan to integrate cloud economics topics from the project into seminars on cloud computing for aspiring college students they regularly give. As part of the project's data management plan, the investigators will also collect, aggregate, and archive real-time market data from public cloud platforms and make it publicly available through a web-based data repository. Finally, the project will emphasize the recruitment of students from under-represented groups.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.

云计算已经成为我们信息经济的基础，为交通、能源、通信、医疗保健、科学和娱乐领域的进步提供了必要的计算能力。早期的云平台很简单，要求用户以固定的每小时价格从少量“按需”服务器中进行选择。然而，现代云平台正在发展成为成熟的市场，根据一系列复杂的合同提供服务器，这些合同使应用程序面临不同形式的风险，例如价格上涨，服务器更换或服务器短缺的风险。 通过提供此类合同，云平台可以提高其基础设施的利用率、收入和能效，而云用户可以通过定制合同来满足应用程序的需求，从而降低成本。 不幸的是，考虑到云合同的复杂性以及风险、应用程序正确性和性能之间的关系，开发风险感知应用程序是一项重大挑战。为了应对这一挑战，该项目将通过调整和扩展经济学和金融学的概念，探索系统支持管理云市场风险的可行性。该项目的一个关键目标是将风险管理提升为一流的系统设计原则。该项目包括多个更广泛的影响。通过实现更好的风险管理，云平台可以显着提高其利用率和能源效率，同时降低用户的成本。该项目将把云经济学和云平台风险管理模块纳入本科生和研究生研讨会，这将使其与操作系统，网络和分布式系统的核心概念相关。研究人员还计划将该项目的云经济学主题整合到他们定期为有抱负的大学生举办的云计算研讨会中。作为该项目数据管理计划的一部分，调查人员还将从公共云平台收集、汇总和存档实时市场数据，并通过基于网络的数据存储库公开提供。 最后，该项目将强调从代表性不足的群体中招募学生。该奖项反映了NSF的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。

项目成果

SpotWeb: Running Latency-sensitive Distributed Web Services on Transient Cloud Servers

• DOI: 10.1145/3307681.3325397
• 发表时间: 2019-06
• 期刊: Proceedings of the 28th International Symposium on High-Performance Parallel and Distributed Computing
• 作者: Ahmed Ali-Eldin;Jonathan Westin;Bin Wang;Prateek Sharma;Prashant J. Shenoy

Understanding Synchronization Costs for Distributed ML on Transient Cloud Resources

• DOI: 10.1109/ic2e.2019.00029
• 发表时间: 2019-06
• 期刊: 2019 IEEE International Conference on Cloud Engineering (IC2E)
• 作者: Pradeep Ambati;David E. Irwin;Prashant J. Shenoy;Lixin Gao;Ahmed Ali-Eldin;Jeannie R. Albrecht

No Reservations: A First Look at Amazon's Reserved Instance Marketplace

• 发表时间: 2020-05
• 期刊: ArXiv
• 作者: Pradeep Ambati;David E. Irwin;Prashant J. Shenoy

Modeling and Analyzing Waiting Policies for Cloud-Enabled Schedulers

• DOI: 10.1109/tpds.2021.3086270
• 发表时间: 2021-12
• 期刊: IEEE Transactions on Parallel and Distributed Systems
• 影响因子: 5.3
• 作者: Pradeep Ambati;Noman Bashir;David E. Irwin;Prashant J. Shenoy

Waiting Game: Optimally Provisioning Fixed Resources for Cloud-Enabled Schedulers

• DOI: 10.1109/sc41405.2020.00071
• 发表时间: 2020-11
• 期刊: SC20: International Conference for High Performance Computing, Networking, Storage and Analysis
• 作者: Pradeep Ambati;Noman Bashir;D. Irwin;Prashant J. Shenoy