Release 389 (7 Jul 2022)#

General#

  • Improve performance of queries involving row type or certain aggregations such as sum, avg, etc. (#12762)

  • Improve performance when spilling to disk is disabled. (#12618)

  • Fix potential incorrect results for queries involving non-equality filters on top of an outer join. (#13109)

  • Fix query failure when no arguments are passed to a table function. (#12951)

  • Fix potential failure when using EXPLAIN with queries involving table functions. (#13106)

  • Fix potential resource leaks when calling long-running regular expression functions. (#13064)

Delta Lake connector#

  • Improve optimized Parquet writer performance for non-structural data types. (#13030)

  • Prevent failure when starting the server if the internal table snapshots cache is disabled. (#13086)

Elasticsearch connector#

  • Add raw_query table function for full query pass-through to the connector. (#12324)

Hive connector#

Iceberg connector#

MongoDB connector#

  • Create a collection when creating a new table. Previously, it was created when the data was written to the table for the first time. (#12892)

Phoenix connector#

  • Add support for Java 17. (#13108)

PostgreSQL connector#

  • Prevent creating a new table with a name longer than the max length. Previously, the name was truncated to the max length. (#12892)

SPI#

  • Remove deprecated version of ConnectorRecordSetProvider#getRecordSet. (#13084)