Release 365 (3 Dec 2021)#
General#
Add support for
TRUNCATE TABLE
. (#8921)Add support for aggregate functions in row pattern recognition context. (#8738)
Add support for time travel queries. (#8773)
Add support for spilling aggregations containing
ORDER BY
orDISTINCT
clauses. (#9723)Add
contains
function to check whether a CIDR contains an IP address. (#9654)Report connector metrics in
EXPLAIN ANALYZE VERBOSE
. (#9858)Report operator input row count distribution in
EXPLAIN ANALYZE VERBOSE
. (#10133)Allow executing
INSERT
orDELETE
statements on tables restricted with a row filter. (#8856)Remove
owner
column from thesystem.metadata.materialized_views
table. (#9961)Remove the
optimizer.iterative-rule-based-column-pruning
config property. The legacy column pruning optimizer is no longer available. (#9564)Improve performance of inequality joins. (#9307)
Improve performance of joins involving a small table on one side. (#9851)
Improve CPU utilization by adjusting
task.concurrency
automatically based on the number of physical cores. (#10088)Make query final query statistics more accurate. (#9888, #9913)
Improve query planning performance for queries containing large
IN
predicates. (#9874)Reduce peak memory usage for queries involving the
rank
,dense_rank
, orrow_number
window functions. (#10056)Fix incorrect results when casting
bigint
values tovarchar(n)
type. (#552)Fix query failure when the
PREPARE
statement is used withDROP
orINSERT
and the table or schema name contains special characters. (#9822)Fix minor memory leak when queries are abandoned during the initial query submission phase. (#9962)
Collect connector metrics after
ConnectorPageSource
is closed. (#9615)
Security#
Allow configuring HTTP proxy for OAuth2 authentication. (#9920, #10069)
Add group-based and owner-based query access rules to file based system access control. (#9811)
Use internal names for discovery client when automatic TLS is enabled for internal communications. This allows
discovery.uri
to be configured using a normal DNS name likehttps://coordinator.trino
and still use automatic TLS certificates. (#9821)Use Kerberos operating system ticket cache if keytab file is not provided to JDBC and CLI for Kerberos authentication. (#8987)
Fix internal communication automatic TLS on Java 17. (#9821)
CLI#
Automatically use HTTPS when port is set to 443. (#8798)
BigQuery connector#
Cassandra connector#
Support reading user defined types in Cassandra. (#147)
ClickHouse connector#
Add support for truncating tables. (#8921)
Fix incorrect query results when query contains predicates on
real
type columns. (#9998)
Druid connector#
Add support for truncating tables. (#8921)
Elasticsearch connector#
Hive connector#
Allow reading empty files of type Parquet, RCFile, SequenceFile. (#9929)
Enable
hive.s3.streaming
by default. (#9715)Improve performance by not generating splits for empty files. (#9929)
Improve performance of decimal
avg
aggregation. (#9738)Improve performance when reading Parquet files with timestamps encoded using
int64
representation. (#9414)Improve dynamic partition pruning efficiency. (#9866, #9869)
Improve query performance on partitioned tables or tables with small files by increasing
hive.split-loader-concurrency
from4
to64
. (#9979)Fix reporting of number of read bytes for tables using
ORC
file format. (#10048)Account for memory used for deleted row information when reading from ACID tables. (#9914, #10070)
Fix
REVOKE GRANT OPTION
to revoke only the grant option instead of revoking the entire privilege. (#10094)Fix bug where incorrect rows were deleted when deleting from a transactional table that has original files (before the first major compaction). (#10095)
Fix delete and update failure when changing a table after a major compaction. (#10120)
Fix incorrect results when decoding decimal values in Parquet reader. (#9971)
Fix
hive.dynamic-filtering.wait-timeout
not having any effect. (#10106)Fix failure when reading Parquet data if column indexes are enabled. (#9890, #10076)
Iceberg connector#
Add support for storing and reading UUID nested in
row
,array
ormap
type. (#9918)Use Iceberg’s
schema.name-mapping.default
table property for scanning files with missing Iceberg IDs. This aligns Trino behavior on migrated files with the Iceberg spec. (#9959)Use ZSTD compression by default. (#10058)
Add read-only security option which can be enabled by setting the configuration
iceberg.security=read-only
. (#9974)Change schema of
$partitions
system table to avoid conflicts when table name contains a column namedrow_count
,file_count
ortotal_size
, or when a column is used for partitioning for part of table data, and it not used for partitioning in some other part of the table data. (#9519, #8729).Improve performance when reading timestamps from Parquet files. (#9414)
Improve query performance for certain queries with complex predicates. (#9309)
Reduce resource consumption and create bigger files when writing to an Iceberg table with partitioning. Bigger files are more efficient to query later. (#9826)
Improve performance for queries on nested data through dereference pushdown. (#8129)
Write correct
file_size_in_bytes
in manifest when creating new ORC files. (#9810)Fix query failures that could appear when reading Parquet files which contained ROW columns that were subject to schema evolution. (#9264)
Fix failure caused by stale metadata in the
rollback_to_snapshot
procedure. (#9921)
Kudu connector#
Avoid scanner time-out issues when reading Kudu tables. (#7250)
MemSQL connector#
Add support for truncating tables. (#8921)
Fix incorrect query results when query contains predicates on
real
type columns. (#9998)
MongoDB connector#
Support connecting to MongoDB clusters via
mongodb.connection-url
config property.mongodb.seeds
andmongodb.credentials
properties are now deprecated. (#9819)
MySQL connector#
Add support for truncating tables. (#8921)
Fix incorrect query results when query contains predicates on
real
type columns. (#9998)
Oracle connector#
Add support for truncating tables. (#8921)
Phoenix connector#
Pinot connector#
Update Pinot connector to be compatible with versions >= 0.8.0 and drop support for older versions. (#9098)
PostgreSQL connector#
Add support for truncating tables. (#8921)
Add experimental support for range predicate pushdown on string columns. It can be enabled by setting the
postgresql.experimental.enable-string-pushdown-with-collate
catalog configuration property or the correspondingenable_string_pushdown_with_collate
session property totrue
. (#9746)
Redshift connector#
Add support for truncating tables. (#8921)
SQL Server connector#
Add support for truncating tables. (#8921)
SPI#
Allow split manager to filter splits based on a predicate not expressible as a
TupleDomain
. (#7608)