15. FileSystem Data Store (HDFS, S3)

The GeoMesa FileSystem datastore (FSDS) provides a cost-effective, performant solution for massive OLAP-style spatial data analytics using frameworks such as Apache Spark. It utilizes modern columnar formats that enable data compression and column-based encodings for efficient OLAP queries. GeoMesa FSDS can run on top of any distributed or local filesystem including Amazon S3, Hadoop HDFS, Google FileStorage, and Azure BlobStore. This enables users to realize cost savings through the utilization of elastic compute resources in lieu of dedicated servers.

The FileSystem datastore can be used to ETL new data into a filesystem or wrap existing file system storage directories using a GeoMesa converter. In both modes the user must provide a defined partition scheme describing the layout of the directory structures.

• 15.1. FileSystem Datastore Architecture
• 15.2. Partition Schemes
• 15.3. Installing GeoMesa FileSystem
• 15.4. FileSystem DataStore Configuration
• 15.5. Using the FileSystem Data Store Programmatically
• 15.6. Using the FileSystem Data Store in GeoServer
• 15.7. FileSystem Command-Line Tools
• 15.8. FileSystem Datastore Example
• 15.9. GeoMesa FileSystem Data Store with Spark SQL