Release 476 (5 Jun 2025)#

General#

Add support for comparing values of geometry type. (#25225)
⚠️ Breaking change: Require JDK 24 to run Trino. (#23498)
Allow configuring query.max-memory-per-node and memory.heap-headroom-per-node relative to maximum heap size. (#25843)
Add feature to deactivate the automated database schema migration for the database backend for resource groups with the property resource-groups.db-migrations-enabled. ({issue)25451)
Make soft memory limit optional in resource groups. (#25916)
Remove the Example HTTP connector from the tar.gz archive and the Docker container. (#25128)
Fix rare bug when server can hang under load. (#25816)
Fix regression introduce in Trino 474 that prevented graceful shutdown from working. (#25690)
Fix potential query failure when the fault_tolerant_execution_runtime_adaptive_partitioning_enabled session property is set to true. (#25870)
Fix failure for queries involving casts with row types. (#25864)
Fix query failures when dynamic catalog names contain mixed case letters. (#25701)
Improve retry logic for S3 operations in file system exchange. (#25908)
Fix query failures when the session catalog or schema names provided by clients contain capital letters. (#25903)

⚠️ Breaking change: Require the --sun-misc-unsafe-memory-access=allow JVM configuration option to run Trino with the connector. (#25669)

Add support for the FOR TIMESTAMP AS OF clause. (#21024)
Add support for user-assigned managed identity authentication for AzureFS. (#23447)
Add signer type support to the native S3 filesystem. (#25820)
Improve compatibility with S3-compliant storage systems. (#25791)
Improve query planning performance. (#24570)
Improve performance when reading tables. (#25826)
Reduce S3 throttling failures. (#25781)
Fix failure when reading variant type column after executing optimize procedure. (#25666)
Fix query failures when attempting to read date columns stored as integer values in Parquet files. (#25667)
Fix failure when querying views without StorageDescriptor on Glue. (#25894)
Fix skipping statistics computation on all columns when only some column types don’t support statistics. (#24487)

Add support for authentication using delegated user credentials with the gsheets.delegated-user-email configuration property. (#25746)

Add support for excluding certain tables from the directory listing cache with the hive.file-status-cache.excluded-tables configuration property. (#25715)
Allow selecting the AWS signing protocol to use when authenticating S3 requests. (#25820)
Improve compatibility with S3-compliant storage systems. (#25791)
Add support for user-assigned managed identity authentication for AzureFS. (#23447)
Improve robustness of the OpenX JSON reader when parsing timestamp values. Previously, only timestamps with a space separator between date and time were supported. (#25792)
Improve metadata reading performance by optimizing directory listing cache behavior. (#25749)
Fix query failures with HIVE_CANNOT_OPEN_SPLIT error when reading ORC files with a large row count. (#25634)
Reduce S3 throttling failures. (#25781)
Fix query failures when attempting to read date columns stored as integer values in Parquet files. (#25667)
Fix failure when querying views without StorageDescriptor on Glue. (#25894)

Add support for user-assigned managed identity authentication for AzureFS. (#23447)
Fix query failures when attempting to read date columns stored as integer values in Parquet files. (#25667)

Add support for user-assigned managed identity authentication for AzureFS. (#23447)
Add signer type support to the native S3 filesystem. (#25820)
Add the added_delete_files_count, existing_delete_files_count, deleted_delete_files_count, and reference_snapshot_id columns to $all_manifests metadata tables. (#25867)
Improve compatibility with S3-compliant storage systems. (#25791)
Show detailed metrics from splits generation in output of EXPLAIN ANALYZE VERBOSE. (#25770)
Add the max_partitions_per_writer catalog session property, which corresponds to the iceberg.max-partitions-per-writer configuration property. (#25662)
Improve query planning performance when reading from materialized views. (#24734)
Prevent rare failure when iceberg.bucket-execution is enabled. (#25125)
Fix query failures with HIVE_CANNOT_OPEN_SPLIT error when reading ORC files with a large row count. (#25634)
Reduce S3 throttling failures. (#25781)
Fix query timeout errors due to concurrent writes on tables with large number of manifest files. (#24751)
Fix query failures when attempting to read date columns stored as integer values in Parquet files. (#25667)
Fix failure when querying views without StorageDescriptor on Glue. (#25894)

⚠️ Breaking change: Remove the kafka-event-listener.client-config-overrides configuration property. To configure the Kafka client for the event listener, specify the configuration in a separate file and set kafka-event-listener.config.resources to the path to the file. (#25553)

Improve performance of selective joins for federated queries. (#25123)
⚠️ Breaking change: Require the --sun-misc-unsafe-memory-access=allow JVM configuration option to run Trino with the connector. (#25669)

Return an estimate of the full data size of the block with getSizeInBytes(). (#25256)