On the Performance of STAC-FastAPI and PgSTAC Using a Cloud-Native Deployment

Storing, processing and delivering geospatial data in large quantities can present a significant challenge due to the sheer amount of available Earth observation satellites, extensively producing copious amounts of data on a daily basis. Given the popularity of cloud-based solutions, it was inevitable that approaches relying on delivering data from public and private clouds have been adopted to facilitate the dissemination of Earth observation data as well. The purpose of this paper is to investigate the performance of recent, cloud native implementations of the data cube concept utilizing technologies developed in the Spatio-Temporal Asset Catalogs (STAC) ecosystem. Specifically, within this paper, we evaluate the performance of STAC-FastAPI and PgSTAC in terms of their scalability and data delivery capabilities by employing the Locust.io framework for testing substantial, concomitant user loads.