S3 Tables Now Supports All Iceberg V3 Data Types, Including Geospatial and Nanosecond Timestamps!

Hi, it's Shii-chan!

This one rounds out type support for data lakes!

What was announced?

According to AWS What's New, Amazon S3 Tables now supports data types defined in the Apache Iceberg Version 3 (V3) specification.

Here's what was added!

With this, S3 Tables support all data types introduced in V3. They already supported the Variant data type, deletion vectors, and row lineage.

You had to encode geospatial coordinates and nanosecond-precision event times in strings or integers. Now you can store them natively.

Geometry and geography columns store points, lines, and polygons natively, so fleet tracking and asset mapping workloads can filter on location at query time
Nanosecond timestamps let telemetry and financial workloads record event times at source precision
Column default values populate a newly added column for existing rows, with no backfill required

AWS says this simplifies data architecture while improving storage efficiency and performance.

S3 Tables provide automatic table maintenance and compaction for Apache Iceberg tables, so tables using V3 data types stay performant and cost effective as data scales
Available in all AWS Regions where S3 Tables are available

See Amazon S3 Tables, Apache Iceberg V3 on AWS, and the AWS Prescriptive Guidance for Iceberg V3 to learn more.

S3 Tables supports all Iceberg V3 data types
geometry, geography, unknown, nanosecond timestamps, and default values are new
Store locations and high-precision times without converting them to strings or integers

This is for data engineers who handle geospatial data or telemetry in Iceberg!