Release 345 (23 Oct 2020)#
General#
Add
concat_ws()
function. (#4680)Add support for
extract()
fortime with time zone
values with precision other than 3. (#5539)Add dynamic filtering support for right joins. (#5555)
Add support for catalog session properties to the file and database backed session property managers. (#5330)
Improve query scalability by increasing the default broadcast join buffer size to 200MB. Broadcast join buffer size can be set using the
sink.max-broadcast-buffer-size
configuration property. (#5551)Improve query performance by allowing larger dynamic filters. (#5529)
Improve performance for join queries where join keys are not of the same type. (#5461)
Improve performance of encrypted spilling. (#5557)
Improve performance of queries that use the
decimal
type. (#5181)
Security#
Add support for JSON Web Key (JWK) to the existing JSON Web Token (JWT) authenticator. This is enabled by setting the
jwt.key-file
configuration property to ahttp
orhttps
url. (#5419)Add column security, column mask and row filter to file-based access controls. (#5460)
Enforce access control for column references in
USING
clause. (#5620)
JDBC driver#
Add
source
parameter for directly setting the source name for a query. (#4739)
Hive connector#
Add support for
INSERT
andDELETE
for ACID tables. (#5402)Apply
hive.domain-compaction-threshold
to dynamic filters. (#5365)Add support for reading Parquet timestamps encoded as microseconds. (#5483)
Improve translation of Hive views. (#4661)
Improve storage caching by better distributing files across workers. (#5621)
Fix disk space accounting for storage caching. (#5621)
Fix failure when reading Parquet
timestamp
columns encoded asint64
. (#5443)
MongoDB connector#
MemSQL connector#
MySQL connector#
Improve performance of queries with aggregations and
LIMIT
clause (but withoutORDER BY
). (#5261)
PostgreSQL connector#
Improve performance of queries with aggregations and
LIMIT
clause (but withoutORDER BY
). (#5261)
Redshift connector#
Add support for setting column comments. (#5397)
SQL Server connector#
Improve performance of queries with aggregations and
LIMIT
clause (but withoutORDER BY
). (#5261)
Thrift connector#
Fix handling of timestamp values. (#5596)