A Quantitative Framework to Inform Cloud Data System Architecture and Services Requirements Based on User Needs and Expected Demand

E Natasha Stavros,Catalina M Oaida,Michelle M Gierach,Jessica Hausman

doi:10.1109/access.2020.3012054

E Natasha Stavros, Catalina M Oaida + Show 2 more

Open Access

https://doi.org/10.1109/access.2020.3012054

Copy DOI

Abstract

The NASA open data policy and the increases in data volume from recent and future missions have resulted in a need to re-evaluate data archive information systems and services. For example, the Surface Water Ocean Topography mission (SWOT) will produce $\sim 15$ TB of data every day that needs to be freely and openly accessible through the NASA Physical Oceanography Distribute Active Archive Center (PO.DAAC). Because of the computational, and distribution challenges associated with making such volumes of Earth Observation data publicly available, the NASA Earth Science Data Information System (ESDIS) is moving their archive systems, like PO.DAAC, to the cloud. To facilitate this migration, system requirements need to accommodate expected use. As such, the objective of this study was to assesses expected user needs and develop a quantitative framework that can be traced to requirements for the cloud-based data archive system architecture and services that enable data discovery, access and utility. We used a two-tier data gathering method that included a short survey of potential users (identified at professional meetings and community listservs) and in-depth interviews. The survey provided a means for scaling findings from the in-depth interviews to a more representative sample, thus enabling assessment of expected impact to various user communities for different services that could be provided. This framework enables traceability for setting priorities for data services development to requirements for a cloud-based data archive system and services that enable data discovery, access and utility.

Highlights

The open data policy by NASA and the increases in data volume from recent and future Earth observing platforms (Table 1) have generated a need to re-evaluate how data is archived in order to provide an optimal design within a sustainable cost model and one which does not jeopardize services for how users discover, access, and use the data
Using the survey to generate anticipated user demographics by application resulted in a queryable table such that most questions could be addressed (Table 2), while providing a means to scale in depth interview results to a representative sample
The method provides an in-depth assessment of user needs based on their current workflows and assesses how these needs may be affected by a future environment on the cloud

Summary

Introduction

The open data policy by NASA and the increases in data volume from recent and future Earth observing platforms (Table 1) have generated a need to re-evaluate how data is archived in order to provide an optimal design within a sustainable cost model and one which does not jeopardize services for how users discover, access, and use the data. Despite this recent request to inform data archive system development, synthesis of input from the larger community will take time, and we have two new missions expected to. Stavros et al.: Quantitative Framework to Inform Cloud Data System Architecture and Services Requirements launch by 2022 that will yield an 89-fold increase in daily data volumes over previous missions. Designing and building the data, archive and distribution systems to support these missions, which must begin in order to be ready and meet mission requirements of public access of verified and validated data post-launch

Objectives

Methods

Results

Discussion

Conclusion