Release 393 (17 Aug 2022)#

General#

Add support for MERGE. (#7933)
Add query state and error code to the timeline log message. (#13698)
Improve performance of highly selective LIMIT queries by making them finish as soon as the required number of rows is produced. (#13616)
Disallow inserting NULL into non-nullable columns.. (#13462)
Fix queries over bucketed tables never completing. (#13655)
Fix potential query failure for GROUP BY queries involving aggregations with ORDER BY and FILTER clauses. (#13627)

Fix potential incorrect rejection of OAuth 2.0 refresh tokens. (#13168)
Fix incorrectly showing outdated OAuth 2.0 refresh tokens in the web UI. (#13168)

Add support for the unsupported-type-handling catalog configuration property. (#13542)
Improve performance for queries with selective joins. (#13334)

Add support for MERGE. (#7933)
Add support for the NOT NULL column constraint. (#13436)
Fix writing incorrect results when the order of partition columns is different from the order in the table definition. (#13505)
Fix failure when reading a table which has partition columns renamed by another engine. (#13521)

Add support for MERGE. (#7933)
Add support for bucket filtering on bucketed columns of float, double, date, list, map and bounded varchar data types. (#13553)
Add exchange.azure.max-error-retries configuration property for the number of retries performed when accessing Azure blob storage. (#13663)
Improve performance of queries with S3 Select pushdown by not utilizing pushdown when it is unnecessary. (#13477)
Reduce Thrift metastore communication overhead when impersonation is enabled. (#13606)
Improve performance when retrieving table statistics from the metastore. (#13488)
Fix error when writing to a table with only date columns while using the Hive metastore. (#13502)
Fix error when reading a Hive view which has a column names with a reserved keyword. (#13450)

Add support for MERGE. (#7933)
Improve performance when filtering on $file_modified_time column. (#13504)
Improve performance of read queries on Iceberg v2 tables with deletion-tracking files. (#13395)
Allow partitioning over columns which use whitespace in their names. (#12226)
Disallow specifying a NOT NULL constraint when adding a new column. Previously, the option was ignored. (#13673)
Fix error when querying tables which are empty and contain no table history. (#13576)
Prevent truncation of the table history in the $snapshots system table by certain DELETE queries. (#12843)
Prevent errors when optimizing an Iceberg table which is empty and contains no table history. (#13582)
Fix incorrect query results when reading from a materialized view that was created on a table which was empty and contained no history. (#13574)

Fix query failure when applying a negative timestamp predicate on the _timestamp column. (#13167)

Prevent renaming a table with a name longer than the max length supported by MongoDB. Previously, the name was truncated to the max length. (#13073)

Improve performance for queries with selective joins. (#13334)
Prevent using schema names or renaming a table with a name which is longer than the max length supported by PostgreSQL. Previously, long names were truncated to the max length. (#13307, #13073)

Improve performance for queries with selective joins. (#13334)
Prevent renaming a table with a name longer than the max length supported by SQL Server. Previously, the name was truncated to the max length. (#13073)

Add @Experimental annotation to designate SPIs that are still under active development. (#13302)
Deprecate io.trino.spi.block.MethodHandleUtil. (#13245)