CC* Data Storage: Supporting Big-Data Edge Computing using Hybrid and Cloud-Native Storage Infrastructure

华盛顿大学(UW)信息技术(UW-IT)和研究副教务长办公室由校园网络基础设施(CC*)奖支持，以创建一个新型的中央数据存储、管理和共享平台。这一研究生态系统的核心承租人是开发可扩展、弹性、低延迟的数据存储框架，以先进的高通量数据科学工具、开源高性能数据管理系统和专业存储网络基础设施工程师的专业知识为后盾，支持广泛的跨学科研究。新的网络基础设施使研究人员能够跨学科和机构边界执行可扩展的数据密集型分析，以便为新发现提供信息，包括材料、粒子物理、环境、医学和健康科学。该项目开发了广泛的数据科学培训计划，为学员提供跨学科的研究经验，并支持合作的少数群体服务机构的数据科学教育和研究，为他们担任不同的教育领导角色做好准备，同时激发更广泛社区对大数据科学的热情。该项目发现并消除了科学领域中常见的数据存储瓶颈，并开发了现代数据存储网络基础设施，该基础设施可将科学分析扩展到本地和现有国家NSF网络基础设施之上的数千个处理器和PB级数据，同时保持与基于云的商业存储的无缝同步。通过在所有三个本地UW园区(西雅图、Bothell、Tacoma)提供可扩展的低延迟数据访问和出口，9.6 PB可用新存储基础设施的部署推动了所有大数据研究学科的发展。新的数据存储生态系统通过以下方式促进研究范式的转变：(1)推动高通量数据分析的发展；(2)通过通用的数据登陆路径鼓励协作的大数据研究；(3)提供高度安全但允许合规使用受监管研究数据的数据管理计划；以及(4)通过汇集显著不同的科学和工程学科来支持融合科学。该奖项反映了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。

The University of Washington (UW) Information Technology (UW-IT) and the Office of Vice Provost for Research are supported by a Campus Cyberinfrastructure (CC*) award to create a novel central data storage, curation, and sharing platform. The core tenant of this research ecosystem is the development of a scalable, elastic, low-latency data storage framework supporting a wide range of interdisciplinary research, bolstered by advanced high-throughput data science tools, an open-source high-performance data management system, and the expertise of professional storage cyberinfrastructure engineers. The new cyberinfrastructure empowers researchers to perform scalable, data-intensive analysis across disciplinary and institutional boundaries to inform new discoveries, including materials, particle physics, environmental, medical, and health sciences. The project develops a broad range of data science training programs to provide interdisciplinary research experiences to the trainees and supports data science education and research at partner minority-serving institutions, preparing them for diverse education leadership roles, and simultaneously inspiring enthusiasm for big-data sciences in the broader community.The project identifies and removes data storage bottlenecks that are common across scientific domains and develops a modern data storage cyberinfrastructure that scales scientific analyses to thousands of processors and petabytes of data locally as well as on top of existing national NSF cyberinfrastructure, while maintaining seamless synchronization with commercial cloud-based storage. The deployment of the 9.6 PB usable new storage infrastructure advances all big data research disciplines by providing scalable low-latency data access and egress across all three local UW campuses (Seattle, Bothell, Tacoma). The new data storage ecosystem facilitates a research paradigm shift by (1) advancing the development of high-throughput data analytics; (2) encouraging collaborative big-data research through a common data landing path; (3) providing a highly secure yet permissive data management plan, to allow compliant use with regulated research data; and (4) supporting convergent science by bringing together substantially different science and engineering disciplines.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.