2 September 2025 - v3.0.0-v5.6.0

Download the Spark Connector JAR.

[v3.0.0-v5.6.0]

Added

• Added handling of unsigned number types
• Added support for casted timestamps in pushdowns
• Added support for partition pruning based on timestamp
• Added support of arrays in WHERE conditions
• Added support of deployment on AWS EMR, AWS EMR Serverless and AWS Sagemaker
• Added support of joining Hydrolix and customer datasets in Databricks
• Added support of Databricks 15.4, 16.4 runtimes

Changed

• Improved retry logic for calls to Hydrolix API
  • Added optional config keys: api_retry_max, api_retry_delay_sec

Fixed

• Fixed NaN ordering in TopN query
• Fixed bug with losing data type information in turbine 5.3+
• Fixed bug with incorrect processing aggregations with unsigned number types
• Fixed bug with filtering by int64/uint32 types issue