As the demands of spatiotemporal studies grow more complex, the need for enhanced computing capabilities has never been clearer. Key tasks, such as managing large datasets, simulating natural processes, and running machine learning algorithms, all call for a strong and adaptive computing environment. Responding to this, our multi-site computing infrastructure is designed to flexibly support a wide array of these research initiatives.
Our advanced system features three distinct computing clusters, together offering 600 nodes that cater to a diverse set of research activities. Our Cloud Platform streamlines the process for team members, allowing for effortless experimentation, data storage, and service deployment. It also integrates resources from leading cloud service providers for easier, more intuitive user access. To address the unique data challenges in spatiotemporal studies, our Big Data Platform is equipped to rapidly process and securely store significant volumes of data. Moreover, our High-Performance Computing Platform accelerates simulations, making it an invaluable tool for various geospatial endeavors.
To maintain the system’s reliability and high performance, we use a monitoring setup powered by Prometheus and Grafana. This gives us real-time insights into resource usage and overall system health, enabling us to ensure that our infrastructure remains responsive to the ever-changing needs of diverse research projects. A visualization of some of our current monitoring metrics can be viewed here.