Release 475 (23 Apr 2025)

General

Add support for the CORRESPONDING clause in set operations. ({issue}25260)
Add support for the AUTO grouping set that includes all non-aggregated columns in the SELECT clause. ({issue}18390)
Publish additional metrics for input tables in event listener. ({issue}25475)
Expose dynamic filter statistics in the QueryCompletedEvent. ({issue}25575)
Improve scalability of inline data encoding in the spooling client protocol. ({issue}25439)
Improve performance of queries involving the exclude_columns table function. ({issue}25117)
Disallow dropping the system catalog. ({issue}24745)
Fix occasional query failures when adaptive planning is enabled. ({issue}25411)
Fix incorrect results when using window functions with DISTINCT. ({issue}25434)
Fix query failures with EXCEEDED_LOCAL_MEMORY_LIMIT errors due to incorrect memory accounting. ({issue}25600)
Properly handle inline session properties for EXPLAIN queries. ({issue}25496)
Add support for exporting OpenTelemetry traces using the HTTP protocol with the tracing.exporter.protocol configuration property set to http/protobuf. ({issue}25573)

Fix incorrect access denial for access control with impersonation when access is granted via the role. ({issue}25166)

Add support for limiting the max parallelism with the bigquery.max-parallelism configuration property. ({issue}25422)
Fix queries getting stuck when reading large tables. ({issue}25423)

Allow cross-region data retrieval when using the S3 native filesystem. ({issue}25200)
Add support for all storage classes when using the S3 native filesystem for writes. ({issue}25435)
Improve performance when filtering on $path, $file_modified_time or $file_size columns. ({issue}25369)
Improve performance of scans on Delta Lake tables with v2 checkpoints. ({issue}25469)

Allow cross-region data retrieval when using the S3 native filesystem. ({issue}25200)
Add support for all storage classes when using the S3 native filesystem for writes. ({issue}25435)
Add support for showing column comments on Hive views. ({issue}23845)
Add support for multiple predicates on partition projection columns with injected types. ({issue}17641)
Fix potential failures or incorrect results when querying partitioned tables using the OpenX JSON SerDe. ({issue}25444)
Ensure Hive metastore locks are released if a failure occurs during lock acquisition. ({issue}25380)
Rename hive.s3.storage-class-filter to hive.s3-glacier-filter to better reflect its purpose. ({issue}25633)
Fix incorrect results when reading timestamp values with leading or trailing spaces using the Regex and OpenX JSON table deserializers. ({issue}25442)
Fix potential performance regression when reading ORC data. ({issue}25617)

Allow cross-region data retrieval when using the S3 native filesystem. ({issue}25200)
Add support for all storage classes when using the S3 native filesystem for writes. ({issue}25435)
Add system.iceberg_tables system table to allow listing only Iceberg tables. ({issue}25136)
Add support for IAM role authentication with the REST catalog. ({issue}25002)
Fix potential failure when queries modify a table concurrently. ({issue}25445)
Add support for returning column statistics for new columns in $partitions system table. ({issue}25532)
Improve the optimize_manifests procedure to produce better organized manifests. ({issue}25378)
Clean up old snapshots when refreshing a materialized view. ({issue}25343)
Set Glue catalog ID when hive.metastore.glue.catalogid is configured. ({issue}25511)
Fix failure when executing migrate on tables partitioned on columns with special characters. ({issue}25106)
Fix OPTIMIZE failures due to commit conflicts with certain DELETE queries. ({issue}25584)
Fix failure when analyzing a table without any snapshots. ({issue}25563)

Fix incorrect results for queries involving LIKE on columns with case-insensitive collations. ({issue}25488)