Skip to main content

Release 466 (27 Nov 2024)

General​

  • Add support for changing the type of row fields when they are in a columns of type map. (#24248)
  • Remove the requirement for a Python runtime on Trino cluster nodes. (#24271)
  • Improve performance of queries involving GROUP BY and joins. (#23812)
  • Improve client protocol throughput by introducing the spooling protocol. (#24214)

Security​

JDBC driver​

  • Improve throughput by automatically using the spooling protocol when it is configured on the Trino cluster, and add the parameter encoding to optionally set the preferred encoding from the JDBC driver. (#24214)
  • Improve decompression performance when running the client with Java 22 or newer. (#24263)
  • Improve performance java.sql.DatabaseMetaData.getTables(). (#24159, #24110)

Server RPM​

  • Remove Python requirement. (#24271)

Docker image​

  • Remove Python runtime and libraries. (#24271)

CLI​

  • Improve throughput by automatically use the spooling protocol when it is configured on the Trino cluster, and add the option --encoding to optionally set the preferred encoding from the CLI. (#24214)
  • Improve decompression performance when running the CLI with Java 22 or newer. (#24263)

BigQuery connector​

  • Add support for LIMIT pushdown. (#23937)

Iceberg connector​

  • Add support for the object store file layout. (#8861)
  • Add support for changing field types inside a map. (#24248)
  • Improve performance of queries with selective joins. (#24277)
  • Fix failure when reading columns containing nested row types that differ from the schema of the underlying Parquet data. (#22922)

Phoenix connector​

  • Improve performance for MERGE statements. (#24075)

SQL Server connector​

  • Rename the sqlserver.experimental.stored-procedure-table-function-enabled configuration property to sqlserver.stored-procedure-table-function-enabled. (#24239)

SPI​

  • Add ConnectorSplit argument to the SystemTable.cursor() method. (#24159)
  • Add support for partial row updates to the ConnectorMetadata.beginMerge() method. (#24075)