Cloud native OpenScienceLabs for HPC : Easing the road to FAIR collaboration and OpenSource

We present a method for publishing high performance compute (HPC) code and results in a scalable, portable and ready-to-use interactive environment in order to enable sharing, collaborating, peer-reviewing and teaching. We show how we utilize cloud native elements such as kubernetes, containerization, automation and webshells to achieve this and demonstrate such an OpenScienceLab for the MAGE (Multiscale Atmosphere Geospace Environment) model, being developed by the recently selected NASA DRIVE Center for Geospace Storms.
We argue that a key factor in the successful design of such an environment is its (cyber)-security, as  these labs require non-trivial compute resources open to a vast audience. Benefits as well as implied costs of different hosting options are discussed, comparing public cloud, hybrid, private cloud and even large desktops.
We encourage HPC centers to test our method using our fully open source blueprints. We hope to thus unburden the research staff and scientists to follow FAIR principles and support open source goals without needing a deep knowledge of cloud computing.