For installation and running instructions, see Get started.
Main changes
SQL features
- SQL commands:
- SQL functions & operators:
- Supports
has_table_privilege()
,has_schema_privilege()
, andhas_any_column_privilege()
. #16674 - Supports
quote_literal()
andquote_nullable()
. #16807 - Supports
pg_get_keywords()
. #17033 - Supports
jsonb_set()
. #17124 - Supports user-defined aggregate functions for embedded Python and JavaScript UDFs. #16874
- Allows the specified time zone of
AT TIME ZONE
to be non-literal. #17395
- Supports
- System catalog:
- Supports
rw_catalog.actor_id_to_ddl
andrw_catalog.fragment_id_to_ddl
. #17229
- Supports
Connectors
- Avro schemas with
"default": "NaN"
and positive and negative infinities, are supported asfloat
anddouble
types. #17309 - Supports ingesting simple
AVRO MAP
types asJSONB
. #16948 - Supports ingesting
avro uuid
types asvarchar
. #17069 - Supports ingesting
avro
with internalRef
types. #17052 - Adds
aws
prefix to AWS related parameters when creating a source or sink. #16671 - Supports using AWS IAM to connect to Amazon MSK. #16625
- Adds
pubsub.parallelism
under theWITH
option for Google PubSub source. #16733 - Supports
INCLUDE TIMESTAMP [AS]
clause for MySQL, PostgreSQL, and MongoDB CDC tables.#16833 - Supports additional metadata columns for CDC tables. #17051
- Automatically maps upstream table schema when creating MySQL and PostgreSQL tables. #16986
- Sets a network timeout for JDBC sink connections. #17244
- Enables sink decouple by default for Kafka, Kinesis, Pulsar, Google Pub/Sub, NATS, MQTT, ClickHouse sinks. #17221
- Supports the
KEY ENCODE
clause when creating a sink. #16377 - Supports
FORMAT PLAIN ENCODE AVRO
for Kafka sinks. #17216 - Supports DynamoDB sink. #16670
- Supports Microsoft SQL Server sinks for self-hosted SQL Server and Azure SQL. #17154
- Supports OpenSearch sink. #16330
- Supports checkpoint decouple for StarRocks sinks. #16816
- Supports checkpoint decouple for Delta Lake sinks. #16777
- Supports sinking serial types. #16969
Cluster configuration changes
- Sets arrangement backfill as the default. #14846
- Supports spill hash join to avoid OOM issue. #17122
- Supports spill hash aggregation for batch queries. #16771
- Changes the algorithm that calculates the reserve memory size. #16992
Fixes
- Improves error message and location of cursor. #16959
- Improves error message when trying to create CDC source with columns. #16636
- Allows
GRANT
andREVOKE
privileges on views. #16699
Full Changelog: v1.9.1...v1.10.0