Release 314 (7 Jun 2019)#
General#
CLI#
Add JSON output format. (#878)
Web UI#
Fix queued queries counter in UI. (#894)
Server RPM#
Change default location of the
http-request.log
to/var/log/presto
. Previously, the log would be located in/var/lib/presto/data/var/log
by default. (#919)
Hive connector#
Fix listing tables and views from Hive 2.3+ Metastore on certain databases, including Derby and Oracle. This fixes
SHOW TABLES
,SHOW VIEWS
and reading frominformation_schema.tables
table. (#833)Fix handling of Avro tables with
avro.schema.url
defined in HiveSERDEPROPERTIES
. (#898)Fix regression that caused ORC bloom filters to be ignored. (#921)
Add support for reading LZ4 and ZSTD compressed Parquet data. (#910)
Add support for writing ZSTD compressed ORC data. (#910)
Add support for configuring ZSTD and LZ4 as default compression methods via the
hive.compression-codec
configuration option. (#910)Do not allow inserting into text format tables that have a header or footer. (#891)
Add
textfile_skip_header_line_count
andtextfile_skip_footer_line_count
table properties for text format tables that specify the number of header and footer lines. (#845)Add
hive.max-splits-per-second
configuration property to allow throttling the split discovery rate, which can reduce load on the file system. (#534)Support overwriting unpartitioned tables for insert queries. (#924)
PostgreSQL connector#
Support PostgreSQL arrays declared using internal type name, for example
_int4
(rather thanint[]
). (#659)
Elasticsearch connector#
Add support for mixed-case field names. (#887)
Base-JDBC connector library#
Allow connectors to customize how they store
NULL
values. (#918)
SPI#
Expose the SQL text of the executed prepared statement to
EventListener
. (#908)Deprecate table layouts for
ConnectorMetadata.makeCompatiblePartitioning()
. (#689)Add support for delete pushdown into connectors via the
ConnectorMetadata.applyDelete()
andConnectorMetadata.executeDelete()
methods. (#689)Allow connectors without distributed tables. (#893)