Release 382 (25 May 2022)#

General#

Add support for fault-tolerant execution with exchange spooling on Google Cloud Storage. (#12360)
Drop support for exchange spooling on S3 with for the legacy schemes s3n:// and s3a://. (#12360)
Improve join performance when one side of the join is small. (#12257)
Fix potential query failures due to EXCEEDED_TASK_DESCRIPTOR_STORAGE_CAPACITY errors with task-based fault-tolerant execution. (#12478)

Add support for using BigQuery’s cached query results. This can be enabled using the bigquery.query-results-cache.enabled configuration property. (#12408)
Support reading wildcard tables. (#4124)

Improve performance of queries that include filters on columns of timestamp with time zone type. (#12007)
Add support for adding columns with ALTER TABLE. (#12371)

Add support for disabling partition caching in the Hive metastore with the hive.metastore-cache.cache-partitions catalog configuration property. (#12343)
Fix potential query failure when metastore caching is enabled. (#12513)
Fix query failure when a transactional table contains a column named operation, originalTransaction, bucket, rowId, row, or currentTransaction. (#12401)
Fix sync_partition_metadata procedure failure when table has a large number of partitions. (#12525)

Support updating Iceberg table partitioning using ALTER TABLE ... SET PROPERTIES. (#12174)
Improves the performance of queries using equality and IN predicates when reading ORC data that contains Bloom filters. (#11732)
Rename the delete_orphan_files table procedure to remove_orphan_files. (#12468)
Improve query performance of reads after DELETE removes all rows from a file. (#12197)

Change ConnectorTableFunction into an interface and add AbstractConnectorTableFunction class as the base implementation of table functions. (#12531)