Cost Optimization on Public Cloud Provider for Big Geospatial Data: A Case Study using Open Street Map

摘要

Big geospatial data is the emerging paradigm for the enormous amount of information made available by the development and widespread use of Geographical Information System (GIS) software. However, this new paradigm presents challenges in data management, which requires tools for large-scale processing, due to the great volumes of data. Spatial Cloud Computing offers facilities to overcome the challenges of a big data environment, providing significant computer power and storage. SpatialHadoop, a fully-fledged MapReduce framework with native support for spatial data, serves as one such tool for large-scale processing. However, in cloud environments, the high cost of processing and system storage in the providers is a central challenge. To address this challenge, this paper presents a cost-efficient method for processing geospatial data in public cloud providers. The data validation software used was Open Street Map (OSM). Test results show that it can optimize the use of computational resources by up to 263% for available SpatialHadoop datasets.