Does Cloud Computing in the Atmospheric Sciences Make Sense? A case study of hybrid cloud computing at NASA Langley Research Center

Wednesday, 17 December 2014
Louis Nguyen1, Thad Chee2, Patrick Minnis1, Douglas Spangenberg2, J Kirk Ayers2, Rabindra Palikonda2, Andrei Vakhnin2, Roger Dubois2 and Patrick R Murphy2, (1)NASA Langley Research Center, Hampton, VA, United States, (2)Science Systems & Applications, Inc., Hampton, VA, United States
The processing, storage and dissemination of satellite cloud and radiation products produced at NASA Langley Research Center are key activities for the Climate Science Branch. A constellation of systems operates in sync to accomplish these goals. Because of the complexity involved with operating such intricate systems, there are both high failure rates and high costs for hardware and system maintenance. Cloud computing has the potential to ameliorate cost and complexity issues. Over time, the cloud computing model has evolved and hybrid systems comprising off-site as well as on-site resources are now common.

Towards our mission of providing the highest quality research products to the widest audience, we have explored the use of the Amazon Web Services (AWS) Cloud and Storage and present a case study of our results and efforts. This project builds upon NASA Langley Cloud and Radiation Group’s experience with operating large and complex computing infrastructures in a reliable and cost effective manner to explore novel ways to leverage cloud computing resources in the atmospheric science environment.

Our case study presents the project requirements and then examines the fit of AWS with the LaRC computing model. We also discuss the evaluation metrics, feasibility, and outcomes and close the case study with the lessons we learned that would apply to others interested in exploring the implementation of the AWS system in their own atmospheric science computing environments.