Effective Software Engineering for Application Development in the Cloud

Cloud-based computing infrastructure provides new capabilities and opportunities to the scientific community through data-sharing, cloud-deployed software, and collaboration tools. From one time use applications, to ongoing long term projects, many have found how these tools can work for them. Additionally, several NOAA datasets already exist within various cloud service providers through the NOAA Big Data Project.

At the Earth System Research Laboratory (ESRL), a team in looking into using cloud computing for a variety of capabilities from modeling, data storage, to web based applications. Research is ongoing into different architectures to help dynamically scale to demand through using kubernetes and containers. Other aspects are investigating leveraging cloud native capabilities like Function as a Service (FaaS) or structuring data storage using different tiers depending on the frequency of access. Through these efforts, we have developed a better understanding of how to make these capabilities work for us, reduce our costs, gain efficiency, and ultimately produce better results for our end users. This presentation will highlight some of these key activities along with techniques to help leverage cloud native services to reduce cost and increase performance four our users users.