Scaling a JupyterHub for Atmospheric Science Research and Education on the NSF Jetstream Cloud

Jupyter notebooks excel at interactive, exploratory scientific programming for researchers and their students. With their mixture of prose, equations, diagrams and interactive code examples, Jupyter notebooks are particularly effective in educational settings and for expository objectives. Their use has become prevalent in many scientific disciplines including atmospheric science. JupyterHub enables specialists to deploy pre-configured notebook servers typically in cloud computing environments. With JupyterHub, users log in to arrive at their own notebook workspace. The advantages of deploying a JupyterHub in a cloud computing environment are numerous. JupyterHub prevents users from having to download and install complex software environments that can be difficult to configure properly. They can be provisioned with computational resources not found in a desktop computing setting and take advantage of high speed networks for processing large datasets. The potential also exists for working on notebooks in a collaborative manner. JupyterHub servers can be accessed from any web browser-enabled device including laptops and tablets. In sum, they greatly improve "time to science" by removing the complexity and tedium required to establish a notebook environment. The challenge at present is deploying these JupyterHub servers in a manner that can "scale" or accommodate many users at once, for example, in a workshop or classroom setting. Unidata is exploring solutions to handle this influx of concurrent JupyterHub users. In particular, several cloud computing technologies have emerged in recent years to address this challenge. Software containerization technologies such as Docker enable experts to install and configure complex software environments on behalf of the user. In addition, Kubernetes is an open-source container-orchestration system for scaling software on cloud computing platforms. This presentation will detail how we leverage the NSF Jetstream Cloud, Docker and Kubernetes to deploy a scalable JupyterHub to improve time to science for atmospheric science researchers and students.