Supported connectors

This topic describes the table types and connectors that Realtime Compute for Apache Flink supports.

Table type definitions

Realtime Compute for Apache Flink uses Flink SQL to define tables that map to upstream and downstream storage. You can also use the Datastream API to connect to storage and perform read and write operations. The following Flink SQL table types are supported:

Source table: The entry point for a data stream. It continuously reads raw event data from external systems, such as Kafka or MySQL CDC.
Dimension table: A reference table used to enrich streaming data. It typically uses a JOIN operation to associate static or infrequently updated dimension information, such as user or product details.
Sink table: The exit point for data processing. It writes computed results, such as aggregated or joined data, to a destination system, such as a data warehouse, database, or message queue.

Relational databases

NoSQL databases

Message queues

Data warehouses

Data lakes

Log and object storage services

Tools and debugging

Key-value and time series databases

General

Connector support details

Connector	Supported Types			Execution mode	API type	Supports updates or deletions to sink table data
Connector	Source table	Dimension table	Sink table	Execution mode	API type	Supports updates or deletions to sink table data
MySQL Note Supports connections to ApsaraDB RDS for MySQL, PolarDB for MySQL, and self-managed MySQL.	√	√	√	Streaming	SQL, DataStream, and YAML for data ingestion	Yes
Hologres	√	√	√	Streaming and batch	SQL, DataStream, and YAML for data ingestion	Yes
ApsaraMQ for Kafka	√	×	√	Streaming	SQL, DataStream, and YAML for data ingestion	Sink tables only support data insertion, not updates or deletions.
MaxCompute	√	√	√	Streaming and batch	SQL and DataStream	Data can be inserted into a sink table, but it cannot be updated or deleted.
Streaming Data Lakehouse Paimon	√	√	√	Streaming and batch	SQL and YAML for data ingestion	Yes
Simple Log Service (SLS)	√	×	√	Streaming	SQL and DataStream	You can only insert data into a sink table. You cannot update or delete data.
StarRocks	√	√	√	Streaming and batch	SQL, DataStream, and YAML for data ingestion	Yes
DataHub	√	×	√	Streaming and batch	SQL and DataStream	The sink table supports inserting data, but not updating or deleting data.
MongoDB	√	√	√	Streaming	SQL and DataStream	Yes
JDBC	√	√	√	Streaming and batch	SQL	Yes
Object Storage Service (OSS)	√	×	√	Streaming and batch	SQL and DataStream	A sink table supports only insert operations. It does not support update or delete operations.
ApsaraDB RDS for MySQL Note The ApsaraDB RDS for MySQL connector will no longer be supported. Use the MySQL connector instead.	×	√	√	Streaming and batch	SQL	Yes
ApsaraMQ for RocketMQ	√	×	√	Streaming	SQL and DataStream	The sink table supports inserts, but not updates or deletes.
Lindorm	×	√	√	Streaming	SQL	Yes
Tair (Redis Open-Source Edition)	×	√	√	Streaming	SQL	Yes
Tair (Enterprise Edition)	×	×	√	Streaming	SQL	Yes
ApsaraDB for HBase	×	√	√	Streaming	SQL	Yes
AnalyticDB for MySQL 3.0	√	√	√	Streaming and batch	SQL	Yes
AnalyticDB for PostgreSQL	×	√	√	Streaming and batch	SQL	Yes
PolarDB for PostgreSQL (compatible with Oracle) 1.0	×	×	√	Streaming and batch	SQL	Yes
Upsert Kafka	√	×	√	Streaming	SQL and YAML for data ingestion	Yes
Elasticsearch	√	√	√	Streaming and batch	SQL and DataStream	Yes
ClickHouse	×	×	√	Streaming and batch	SQL	Yes
Tablestore (OTS)	√	√	√	Streaming	SQL	Yes
SelectDB	×	×	√	Streaming and batch	DataStream and SQL	Yes
Time Series Database for InfluxDB	×	×	√	Streaming	SQL	No
Postgres CDC (public preview)	√	×	×	Streaming	SQL	N/A
OceanBase (public preview)	√	√	√	Streaming and batch	SQL	Yes
Iceberg	√	×	√	Streaming and batch	SQL	Yes
Generating simulated data with Faker	√	√	×	Streaming and batch	SQL	N/A
Blackhole	×	×	√	Streaming and batch	SQL	Yes
Datagen	√	×	×	Streaming and batch	SQL	N/A
Print	×	×	√	Streaming and batch	SQL and YAML for data ingestion	Yes
Hudi (being deprecated)	√	×	√	Streaming and batch	SQL and DataStream	Yes