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 BYand joins. (#23812)Improve client protocol throughput by introducing the spooling protocol. (#24214)
JDBC driver#
Improve throughput by automatically using the spooling protocol when it is configured on the Trino cluster, and add the parameter
encodingto 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
--encodingto 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
LIMITpushdown. (#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
MERGEstatements. (#24075)
SQL Server connector#
Rename the
sqlserver.experimental.stored-procedure-table-function-enabledconfiguration property tosqlserver.stored-procedure-table-function-enabled. (#24239)