CC* Networking Infrastructure: Advanced Network For Research at UNC Charlotte

来自北卡罗来纳大学夏洛特分校的教师研究人员、研究计算和网络专家组成的不同团队正在协作开发和部署一个先进的网络，用于整个校园的研究。该网络极大地改善了研究人员、科学仪器、可视化工作站、高性能计算(HPC)基础设施和外部合作者之间的数字通信。通过访问高级网络，可以更快地访问现有研究工作流程，并支持团队以前无法使用的新研究流程。例如，来自多台DNA测序仪的海量数据集被传输到功能强大的HPC集群，从而消除了数据分析瓶颈。将运动分析和人工智能应用于未来工作研究、自然语言处理和网络物理系统的研究也得到了支持。应用包括对公共交通中携带传染性微生物的气流进行流体动力学模拟。北卡罗来纳大学夏洛特分校和其他大学之间的研究合作通过加快研究成果的获取和共享而得到加强。来自不同背景的学生，包括北卡罗来纳大学夏洛特分校来自代表性不足群体的大量群体和第一代大学生，将在他们的课程和研究项目中使用这种最先进的网络基础设施。该设计是一个校园范围的基于100 Gb光纤的网络，带有Science DMZ，包括一个数据传输节点，支持将数据流与大学业务的日常流量分开。该网络实施为主干-枝叶架构，两个主干(64x100 GbE)通过MLAG通过100 Gb链路相互连接。一个主干放在HPC数据中心，另一个放在单独的园区数据中心。该网络具有6个枝叶(32x100 GbE)，支持学术部门所在大楼和HPC基础设施之间的连接。性能通过PerfSonar节点进行监控和调整。安全通过Zeek(40 Gb/S)节点进行监控。网络和科学非军事区是在研究人员和网络架构之间的合作中进行调整和优化的。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。

A diverse team of faculty researchers, research computing, and networking experts from across UNC Charlotte are collaborating to develop and deploy an advanced network for research across the campus. This network dramatically improves digital communication between researchers, scientific instrumentation, visualization workstations, high performance computing (HPC) infrastructure and external collaborators. Access to the advanced network enables higher speed to existing research workflows and allows for new research processes previously unavailable to the team. For example, massive datasets from multiple DNA sequencing instruments are streamed to powerful HPC clusters thus removing data analysis bottlenecks. Research applying motion analysis and artificial intelligence to Future of Work research, natural language processing and cyber physical systems are also supported. Applications include fluid dynamics simulation of airflow, which carries infectious microbes, in public transportation. Research collaborations between UNC Charlotte and other universities is enhanced by speeding acquisition and sharing of research results. Students from varied backgrounds, including UNC Charlotte’s large cohorts from underrepresented groups and first generation college students will work with this state-of-the-art cyberinfrastructure in their coursework and research projects. The design is a campus wide 100Gb fibre-based network with Science DMZ including a Data Transfer Node enabling data flows separate from the day-to-day traffic of University business. The network is implemented as a spine-leaf architecture with two spines (64x100GbE) interconnected to each other with 100Gb links by MLAG. One spine is placed in the HPC Data Center and the other is placed in a separate campus Data Center. The network has 6 Leafs (32x100 GbE) supporting connectivity between buildings housing academic departments and HPC infrastructure. Performance is monitored and tuned with PerfSonar nodes. Security is monitored through Zeek (40Gb/s) nodes. The network and Science DMZ are tuned and optimized in collaboration between researchers and network architects.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.