Release 316 (8 Jul 2019)#
General#
Fix
date_format
function failure when format string contains non-ASCII characters. (#1056)Improve performance of queries using
UNNEST
. (#901)Improve error message when statement parsing fails. (#1042)
CLI#
Fix refresh of completion cache when catalog or schema is changed. (#1016)
Allow reading password from console when stdout is a pipe. (#982)
Hive connector#
Acquire S3 credentials from the default AWS locations if not configured explicitly. (#741)
Only allow using roles and grants with SQL standard based authorization. (#972)
Add support for
CSV
file format. (#920)Support reading from and writing to Hadoop encryption zones (Hadoop KMS). (#997)
Collect column statistics on write by default. This can be disabled using the
hive.collect-column-statistics-on-write
configuration property or thecollect_column_statistics_on_write
session property. (#981)Eliminate unused idle threads when using the metastore cache. (#1061)
PostgreSQL connector#
Add support for columns of type
UUID
. (#1011)Export JMX statistics for various JDBC and connector operations. (#906).
MySQL connector#
Export JMX statistics for various JDBC and connector operations. (#906).
Redshift connector#
Export JMX statistics for various JDBC and connector operations. (#906).
SQL Server connector#
Export JMX statistics for various JDBC and connector operations. (#906).
TPC-H connector#
Fix
SHOW TABLES
failure when used with a hidden schema. (#1005)
TPC-DS connector#
Fix
SHOW TABLES
failure when used with a hidden schema. (#1005)
SPI#
Add support for pushing simple column and row field reference expressions into connectors via the
ConnectorMetadata.applyProjection()
method. (#676)