SedonaDB is a full-fledged spatial data lake + data warehouse system for processing, querying and storing spatial data at any scale. It provides Spatial SQL, Scala, Java, and Python interfaces for you to drive insights from your data.
- LakeHouse Architecture: SedonaDB employs Havasu which is a spatial table format built on top of Apache Iceberg. Havasu supports ACID transactions, schema evolution, and time travel. Havasu stores spatial data in Apache Parquet format on S3, which is very cost effective and highly decoupled with computation.
- Spatial Type Support: SedonaDB tables accord first-class status to both geometry and raster data. Users can effortlessly store and query either type of data using Sedona SQL within the same notebook, avoiding the frustrating and slow experience typically associated with toggling between different traditional software platforms.
- Fast Spatial Query: SedonaDB adeptly comprehends and optimizes spatial queries, thereby achieving remarkably fast query speeds. Havasu has the capability to reorganize and index spatial data based on its spatial proximity. The query engine of SedonaDB employs highly efficient distributed spatial query algorithms, ensuring optimized performance.
- Rich Spatial Functions: SedonaDB offers a significantly broader range of functions and markedly higher speed compared to both Apache Sedona and Apache Spark, yet it maintains support for the existing public APIs found in Apache Sedona and Apache Spark.
Last update: October 17, 2023 23:45:31