12.27. Release 308
General Changes
- Fix a regression that prevented the server from starting on Java 9+. (#610)
- Fix correctness issue for queries involving
FULL OUTER JOIN and coalesce. (#622)
Security Changes
- Add authorization for listing table columns. (#507)
CLI Changes
- Add option for specifying Kerberos service principal pattern. (#597)
JDBC Driver Changes
- Correctly report precision and column display size in
ResultSetMetaData
for char and varchar columns. (#615)
- Add option for specifying Kerberos service principal pattern. (#597)
Hive Connector Changes
- Fix regression that could cause queries to fail with
Query can potentially
read more than X partitions error. (#619)
- Improve ORC read performance significantly. For TPC-DS, this saves about 9.5% of
total CPU when running over gzip-compressed data. (#555)
- Require access to a table (any privilege) in order to list the columns. (#507)
- Add directory listing cache for specific tables. The list of tables is specified
using the
hive.file-status-cache-tables configuration property. (#343)
MySQL Connector Changes
- Fix
ALTER TABLE ... RENAME TO ... statement. (#586)
- Push simple
LIMIT queries into the external database. (#589)
PostgreSQL Connector Changes
- Push simple
LIMIT queries into the external database. (#589)
Redshift Connector Changes
- Push simple
LIMIT queries into the external database. (#589)
SQL Server Connector Changes
- Fix writing
varchar values with non-Latin characters in CREATE TABLE AS. (#573)
- Support writing
varchar and char values with length longer than 4000
characters in CREATE TABLE AS. (#573)
- Support writing
boolean values in CREATE TABLE AS. (#573)
- Push simple
LIMIT queries into the external database. (#589)
Elasticsearch Connector Changes
- Add support for Search Guard in Elasticsearch connector. Please refer to Elasticsearch Connector
for the relevant configuration properties. (#438)
