Release 453 (25 Jul 2024)#

General#

  • Improve accuracy of the cosine_distance() function. (#22761)

  • Improve performance of non-equality joins. (#22521)

  • Improve performance for column masking with Open Policy Agent access control. (#21359)

  • Fix incorrect evaluation of repeated non-deterministic functions. (#22683)

  • Fix potential failure for queries involving GROUP BY, UNNEST, and filters over expressions that may produce an error for certain inputs. (#22731)

  • Fix planning failure for queries with a filter on an aggregation. (#22716)

  • Fix planning failure for queries involving multiple aggregations and CASE expressions. (#22806)

  • Fix optimizer timeout for certain queries involving aggregations and CASE expressions. (#22813)

Security#

  • Add support for IF EXISTS to DROP ROLE. (#21985)

JDBC driver#

  • Add support for using certificates from the operating system keystore. (#22341)

  • Add support for setting the default SQL PATH. (#22703)

  • Allow Trino host URI specification without port for the default ports 80 for HTTP and 443 for HTTPS. (#22724)

CLI#

  • Add support for using certificates from the operating system keystore. (#22341)

  • Add support for setting the default SQL PATH. (#22703)

  • Allow Trino host URI specification without port for the default ports 80 for HTTP and 443 for HTTPS. (#22724)

BigQuery connector#

  • Improve performance when querying information schema. (#22770)

Cassandra connector#

  • Add support for the execute procedure. (#22556)

ClickHouse connector#

  • Add support for the execute procedure. (#22556)

Delta Lake connector#

  • Add support for reading partition columns whose type changed via type widening. (#22433)

  • Add support for authenticating with Glue with a Kubernetes service account. This can be enabled via the hive.metastore.glue.use-web-identity-token-credentials-provider configuration property. (#15267)

  • Fix failure when executing the VACUUM procedure on tables without old transaction logs. (#22816)

Druid connector#

  • Add support for the execute procedure. (#22556)

Exasol connector#

  • Add support for the execute procedure. (#22556)

Hive connector#

  • Add support for authenticating with Glue with a Kubernetes service account. This can be enabled via the hive.metastore.glue.use-web-identity-token-credentials-provider configuration property. (#15267)

  • Fix failure to read Hive tables migrated to Iceberg with Apache Spark. (#11338)

  • Fix failure for CREATE FUNCTION with SQL routine storage in Glue when hive.metastore.glue.catalogid is set. (#22813)

Hudi connector#

  • Add support for authenticating with Glue with a Kubernetes service account. This can be enabled via the hive.metastore.glue.use-web-identity-token-credentials-provider configuration property. (#15267)

Iceberg connector#

  • ⚠️ Breaking change: Change the schema version for the JDBC catalog database to V1. The previous value can be restored by setting the iceberg.jdbc-catalog.schema-version configuration property to V0. (#22576)

  • Add support for views with the JDBC catalog. Requires an upgrade of the schema for the JDBC catalog database to V1. (#22576)

  • Add support for specifying on which schemas to enforce the presence of a partition filter in queries. This can be configured query-partition-filter-required-schemas property. (#22540)

  • Add support for authenticating with Glue with a Kubernetes service account. This can be enabled via the hive.metastore.glue.use-web-identity-token-credentials-provider configuration property. (#15267)

  • Fix failure when executing DROP SCHEMA ... CASCADE using the REST catalog with Iceberg views. (#22758)

Ignite connector#

  • Add support for the execute procedure. (#22556)

MariaDB connector#

  • Add support for the execute procedure. (#22556)

MySQL connector#

  • Add support for the execute procedure. (#22556)

Oracle connector#

  • Add support for the execute procedure. (#22556)

Phoenix connector#

  • Add support for the execute procedure. (#22556)

PostgreSQL connector#

  • Add support for reading the vector type on pgvector. (#22630)

  • Add support for the execute procedure. (#22556)

Redshift connector#

  • Add support for the execute procedure. (#22556)

SingleStore connector#

  • Add support for the execute procedure. (#22556)

Snowflake connector#

  • Add support for the execute procedure. (#22556)

SQL Server connector#

  • Add support for the execute procedure. (#22556)

SPI#

  • Add SystemAccessControl.getColumnMasks as replacement for the deprecated SystemAccessControl.getColumnMask. (#21997)