Release 0.225

    • Fix leak in operator peak memory computations. (#13210)

    • Allow overriding session zone sent by clients. ()

    • Throw an exception when a spatial partition is made with 0 rows.

    • Add field containing statistics estimates for input tables to the QueryInputMetadata in QueryCompletedEvent for join queries. (#12808)

    JDBC Changes

    • Match schema and table names case insensitively. This behavior can be enabled by setting the catalog configuration option to true.
    • Display tasks per stage on the query page.

    Elasticsearch Connector Changes

    • Fix and support nested fields in Elasticsearch Connector. ()

    • Improve performance of file reads on S3. (#13222)

    • Add ability to use client-provided OAuth token to access Google Cloud Storage. This can be configured using the configuration property hive.gcs.use-access-token=false.

    Raptor Changes

    • Return an error when creating tables with unsupported types like ROW type. (#13209)

    • Remove legacy ORC writer. Configuration property storage.orc.optimized-writer-stage is enabled by default. Option DISABLED is removed.

    • Add TableHandle parameter to ConnectorPageSourceProvider#createPageSource method. This gives connector access to ConnectorTableLayoutHandle during execution.

    • Add columnHandles parameter to method. The new parameter allows connectors to prune statistics to the desired list of columns and subfields and fixes compatibility issue between subfield pruning and CBO. ()

    Verifier Changes

    • Add support for overriding session properties for all queries in a suite.

    • Add cluster type and retryable information for QueryFailure.

    • Add final query failure information to Verifier output event.